As more sex workers and activists, organizers, and protesters (AOP) move online due to COVID-19, the sex working community and organizing efforts are being disrupted through legislative efforts to increase surveillance and platform liability. Sex worker contributions to movement work are often erased, despite the fact that a significant amount of unpaid activism work (specific to sex work or otherwise) is funded by activists’ direct labor in the sex trades. This research aims to gain a better understanding of the ways that platforms’ responses to Section 230 carve-outs impact content moderation, and threaten free speech and human rights for those who trade sex and/or work in movement spaces. In this sex worker-led study, Hacking//Hustling used a participatory action research model to gather quantitative and qualitative data to study the impact of content moderation on sex workers and AOP (n=262) after the uprisings against state-sanctioned police violence and police murder of Black people. The results of our survey indicate that sex workers and AOP have noticed significant changes in content moderation tactics aiding in the disruption of movement work, the flow of capital, and further chilling speech. The negative impact of content moderation experienced by people who identified as both sex workers and AOP was significantly compounded.
Key Words: Sex Work, Prostitution, Content Moderation, Section 230, Tech, Public Health, Platform Policing, Censorship, Community Organizing, Activist, Platform Liability, Free Speech, First Ammendment
Hacking//Hustling is a collective of sex workers, survivors, and accomplices working at the intersection of tech and social justice to interrupt state surveillance and violence facilitated by technology. Hacking//Hustling works to redefine technologies to uplift survival strategies that build safety without prisons or policing. It is a space for digital rights advocates, journalists, and allied communities to learn from sex workers. In an effort to fill in gaps of knowledge the academy and policy makers neglect, we employ feminist data collection strategies and participatory community-based research models to assess the needs of our community. This collective was formed with the belief that sex workers are the experts of their own experience, and that an Internet that is safer for sex workers is an Internet that is safer for everyone.
Much love, appreciation, and care to our resilient sex working/trading community for sharing your insights and analysis, your organizing and your survival strategies. We are grateful to everyone who took the time to complete this survey, offered their time and labor for peer review, to JB Brager for their beautiful illustrations, and to our graphic designer, Livia Foldes, who volunteered her labor and genius. This report was made possible with funding in part from Hacking//Hustling, but largely with funding from our researchers’ direct labor in the sex trades or access to employment in formal economies.
Danielle Blunt (she/her) is a NYC-based Dominatrix, a full-spectrum doula, and chronically-ill sex worker rights activist. She has her Masters of Public Health and researches the intersection of public health, sex work, and equitable access to technology in populations vulnerable to state and platform policing. Blunt is on the advisory board of Berkman Klein’s Initiative for a Representative First Amendment (IfRFA) and the Surveillance Technology Oversight Project in NYC. She enjoys redistributing resources from institutions, watching her community thrive, and making men cry. DanielleBlunt@protonmail.ch
Ariel Wolf (she/her) is a writer, researcher, and former sex worker from New York City. She previously served as the community organizer for the Red Umbrella Project, a nonprofit organization that amplified the voices of those in the sex trades, and as a research assistant for the Center for Court Innovation, where she has worked on studies on sex work, human trafficking, gun violence, and procedural justice. Her first solo academic paper entitled “Stigma in the Sex Trades” was published in the Journal of Sexuality and Relationship Therapy. ArielHWolf@gmail.com
Emily Coombes (they/them) is a scholar-organizer based in Las Vegas. Their work investigates the impacts of increasing surveillance and changing digital landscapes on political mobilization. A PhD student at the University of Nevada, Las Vegas, their research focuses on the 2018 sex worker-led hashtag campaign #LetUsSurvive #SurvivorsAgainstSESTA launched against FOSTA/SESTA. Emily previously worked as the North American Regional Correspondent to the Global Network of Sex Work Projects (NSWP) and currently serves as the Resident Movement Scholar for Hacking//Hustling. firstname.lastname@example.org
Shanelle Mullin (she/her) is a jill-of-all-trades marketer with a 13-year background in growth and tech. By day, she leads conversion rate optimization at the fastest growing SaaS company in history. By night, she’s a freelance writer in the marketing and tech space. She was recently named one of the world’s top experts on conversion rate optimization. She studies web psychology, online user behaviour, and human rights in tech. email@example.com
Table of Contents
- Section 230, Free Speech Online, and Social Media Censorship
- Key Findings
- Sex Workers vs. Activists, Organizers, Protestors (AOPs) vs. Both
- Sex Work and COVID-19
- Sex Work and Black Lives Matter
- Chilled Speech
- Organizing Under a Shadowban & Disruption of Movement Spaces
- Financial Technology
- Discussions and Recommendations
- Platform Recommendations
- Policy Recommendations
- Further Research
- Important Terms and Concepts
As we look to more censorship, surveillance, and carve-outs to Section 230 (§ 230) on the horizon, we believe that it is important to understand how these carve-outs impact not just social and financial platforms and their content moderation decisions, but the humans who rely on those platforms every day. The COVID-19 pandemic, ongoing uprisings against police violence, and the increasingly rapid flow of data between state and private actors highlight how state and platform policing impacts communities both online and in the streets.
Our peer-led landscape analysis of content moderation and changes in online experiences between May 25th and July 10th, during the uprisings against the most recent police murders of Black people, provides analytical insight to how content moderation systems and amendments to § 230 disrupt freedom of speech—and human rights—in sex workers’ and AOPs’ digital lives. It is important to note that algorithms are constantly evolving and being manually updated, but the data in this report mirrors the trends found in our previous study, Erased.
The ways that sex work and organizing are policed on the streets through racist, transphobic policing tactics and use of condoms as evidence, paralell the inequitable ways these communities are policed online: content moderation, shadowbanning, and denial of access to financial technologies. While this report begins to touch on how content moderation practices, deplatforming, and online surveillance can follow people offline, this report just brushes the surface of the extent of how surveillance technology impacts communities vulnerable to policing and in street-based economies.
We want to be clear that this is an academic paper, which may not be accessible to all members of our community. While we worked to make this paper as accessible to community as possible while still maintaining academic rigor, we acknowledge that the academy—and much of the tech industry— was created by a social class who have the time, leisure, and money to pursue higher education instead of working. So if this paper is inaccessible and overwhelming to you, it is not your fault. In an attempt to make this research more accessible, we have provided a glossary of important terms and concepts at the end of the report. We will also be presenting our findings in a live presentation, where community can ask questions and engage the researchers in conversation. A recording of this video will be archived on our YouTube page (with a transcript).
Our current research explores the intersection of sex workers’ and AOPs’ online experiences, and seeks to better understand how content moderation impacts their ability to work and organize, both online and offline. This research highlights the harm laws like FOSTA and the EARN IT Act can cause to communities vulnerable to surveillance and policing, including the victims and survivors that many of these bills purport to protect. In this project, we explore how different communities experience content moderation online. This report serves as an extension of Erased, our study on FOSTA, and adds to the small body of research that focuses on the human impact of § 230 carve-outs and platform content moderation decisions.
Content moderation, censorship, and shadowbanning facilitate sex worker erasure and normalize the digital and physical opression of sex working and AOP communities. Sex workers are disproportionately losing access to social media platforms, having bank accounts seized, being banned from major payment processers, and being used as test subjects for facial recognition databases. These are forms of structural violence that predominantly impact populations already vulnerable to state and platform policing’s access to resources, community, and harm reduction materials. This research shows how communities are negatively impacted by content moderation practices, and how surveillance technologies disrupt their ability to both earn an income and organize.
Section 230, Free Speech Online, and Social Media Censorship
In 2018, we saw the first substantive and successful attempts to dismantle § 230 with Public Law 115-164, better known as a combination of FOSTA H.R. 1865 (“Allow States and Victims to Fight Online Sex Trafficking Act”) and SESTA S.B. 1693 (“Stop Enabling Sex Trafficking Act”). FOSTA broadly expanded civil and criminal liability for websites with user-generated content, including: Twitter, Instagram, and many sites that sex workers advertised their services on. FOSTA is just one part of a larger whorephobic ecosystem that facilitates the erasure of sex workers from online spaces. FOSTA follows a broader trend of sex workers losing access to online spaces, such as with the FBI raidings of RentBoy, Backpage, and Eros.
Sex workers have been experiencing the collateral damages of private companies trying to demonstrate due diligence and mitigate other forms of liability. In our previous study, Erased, Hacking//Hustling’s sex worker-, peer-led research team found: 94% of online respondents say they advertise sex work-related services using online public platforms and social media; 99% do not feel safer because of FOSTA; 72.45% say FOSTA plays a role in their increased economic instability; 33.8% report an increase of violence from clients; 80.61% are now facing increased difficulties advertising their services; and 21% are not able to access online harm reduction anymore. In Erased, we showed how FOSTA encourages platforms to contribute to the silencing and speech chilling of survivors, sex workers, and sex working survivors through erasing sex workers from the Internet. There is already tremendous fear in the community as sex workers try to comply with and work around platform rules that are often opaque and enforced differently for different people. These findings are also confirmed in the findings of COYOTE-RI’s 2018 survey on the impacts of FOSTA.
This year, EARN IT S. 3398 (“Eliminating Abusive and Rampant Neglect of Interactive Technologies Act of 2020”) was introduced with bi-partisan support. The EARN IT Act seeks to further compromise Internet freedom, and digital and human rights—for everyone—under the guise of preventing child sexual abuse material (CSAM). The EARN IT Act makes the incorrect assumption that sex workers and survivors are two distinct communities when our lived experiences are often much more complicated and our needs are not in opposition. EARN IT would increase civil and criminal liability for platforms while only providing legal recourse for very few survivors of CSAM. Worse, it would harm many sex workers, survivors, and sex working survivors while providing no meaningful resources to actually stop child sexual abuse.
Technology has historically been used by the U.S. Government to repress and silence movement-building efforts, and magnify systems of oppression and violence that criminalize and police communities and how they support themselves. With bills like the EARN IT Act, PACT Act, and Lawful Access to Encrypted Data Act on the horizon as well as the US Agency for Global Media’s hostile takeover of the Open Technology Fund, we believe that now is the time to better understand the impact of content moderation on human rights and movement work. As the pandemic necessitates more online interactions and decreased access to public spaces, we are moving into an ecosystem of increased government-sanctioned surveillance and censorship. We need to question tech’s primary and flawed solution to mitigating liability—content moderation—and the role it plays in our lives, our communities and our organizing.
The Communication Decency Act (CDA) of 1996 was Congress’ first significant attempt to regulate online content. The anti-indecency provision of the CDA was struck down in Reno v. ACLU, but what remained was § 230. Section 230 states: “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.”
Section 230 was created in response to conflicting case law in the early 1990s that established that websites that host user-generated content would be treated like bookstores. Thus, websites would not be legally liable for the content they host unless they moderated their content in any way. If the websites were moderated, those sites would be treated like traditional publishers and held legally responsible for defamatory or obscene content posted by their users.
Perversely, those early cases incentivized website owners to not moderate user-generated content because such content moderation would increase their legal liability. While section 230 has been largely credited for creating free speech protections online, what it actually did was protect content moderation. After § 230, websites could moderate user-generated content without being treated as publishers under the law, meaning they would not be punished if they missed defamatory or obscene user-generated content while moderating. This legal protection encouraged content moderation, but not suppression of user speech—enabling innovation and open discourse in online spaces.
On April 11th, 2018, FOSTA was signed into law with bi-partisan support. This bill was sponsored by Senator Rob Portman with lobbying support from organizations, including: The New Jersey Coalition Against Human Trafficking, ECPAT Omtermatopma, Operation Texas Shield, and Faith & Freedom Coalition. Corporations such as 21st Century Fox and Oracle Corporation also voiced their support for FOSTA. The Internet Association—which represents companies such as Facebook, Google and Microsoft—initially voiced opposition to the bill, which they later withdrew after minor changes to the wording of one section. In the end, the passing of FOSTA was supported by Big Tech, who benefited from the pressure it would put on smaller competitors to shutter their sites for fear of legal liability, and endorsed by multiple celebrities, including Amy Schumer and Seth Myers.
One part of FOSTA was the first substantive amendment to §230. FOSTA’s stated Congressional purpose was to make it “easier for prosecutors and others to hold websites criminally and civilly liable when those websites are used to facilitate prostitution or sex trafficking.” What the law has actually done is increased Internet platform liability and put pressure on them to censor their users and push communities into increased financial insecurity, housing instability, and exposure to violence.
Congress ignored warnings from sex workers, survivors of trafficking, and sex working survivors on what the human impact of this bill would be. Subsequent community-based research proves that this bill has not only done nothing to address human trafficking,, but has pushed communities into increased vulnerability as well. Now, amidst a pandemic, when online communication is particularly important, legislators are attempting to pass more bills that amend § 230, threatening to destroy the affordances for open discourse that § 230 facilitated.
It is important to note that U.S. Internet legislation is not contained to the U.S. It has international impact as many companies host content in the U.S.
On December 17th, 2019, known as The International Day to End Violence Against Sex Workers, Representatives Ro Khanna and Barbara Lee and Senators Elizabeth Warren and Ron Wyden introduced the SAFE Sex Workers Study Act. This bill is the first of its kind, which would require Congress to study the health and safety of sex workers and the associated impacts of FOSTA. Congress has still not progressed this bill, despite acknowledgement from electeds that FOSTA has caused harm.
Congress is not: supporting the SAFE Sex Workers Study Act; supporting Medicaid for all, which would provide victims of child sexual abuse with more options for medical health care, mental health care, and spaces to begin healing from trauma; or supporting comprehensive sex education. Instead, Congress is fast-tracking the EARN IT Act to create further carve-outs to § 230 during a pandemic (without putting resources toward actually ending child sexual abuse).
The Patriot Act—passed in 2001, just after 9/11—increased the surveillance powers of the U.S. Government. One of the stated goals of the Patriot Act was to make it easier for state and federal agencies to share information. We see parallels in anti-terrorism and anti-trafficking rhetoric: the narrative of fear creates the need for increased surveillance. Once the need is gone, the increased surveillance methods stay, impacting everyone, but especially Black and Muslim sex workers who experience multiple intersecting forms of surveillance and policing.
Kendra Albert, a technology lawyer, describes this process as, “Data that’s collected across multiple methods of surveillance and putting it together to gain more information about the lives of individual people.” In recent history, we’ve seen this violence coming through via amendments to CDA 230, content moderation, and threats to encryption.
FOSTA and the EARN IT Act are part of a long history of the state stoking fear for political gain, leading to legislation that erodes privacy and free speech for all and increases the surveillance of already heavily criminalized and policed communities. Through invoking “white slavery” myths, the state is able to rationalize mass surveillance policies and censorship. When a community is vilified by the media, it creates an environment where surveillance and policing seem like the only options.
Who Makes Tech & Inherited Biases
The majority of the digital technologies we rely on for daily communications are created by those already in power: wealthy, white, cishetero men. Many of the digital technologies that we use are created as part of a broader system of white supremacy and technocapitalism. These technologies reflect the biases of their creators, and can serve as extensions of the carceral state when they are used to deepen oppression and surveil communities. Many of these processes rely on opaque algorithm-driven software. While representatives of large tech corporations insist the algorithms and processes driving their content moderation are neutral, it has been shown time and time again that they are not.
Joy Buolamwini and the work of the Algorithmic Justice League demonstrate how, in a time when AI is increasingly governing our everyday lives, machine learning is encoded with racial and gender biases. Similarly, Safiya Umoja Noble has explained the oppressive impact of algorithmic biases, especially within search engines, in her book Algorithms of Oppression. This reality leaves us vulnerable to a digital world where racism, sexism, and transphobia are quite literally coded into the services, platforms, and automated processes that we rely on every day to live and work.
Content moderation, “algorithms of oppression”, and surveillance tech have been weaponized against communities of color in a wide range of ways, both implicitly and explicitly. For example, facial recognition technologies, which are becoming more widely used, lead to the disproportionate incarceration of Black people and communities of color.
The increasing severity of content moderation online is a key component of the growing surveillance and silencing of communities already vulnerable to state and platform policing. The online technologies that are actively policing and criminalizing sex work contribute to a broader system of state-corporate funded surveillance. The increased and opaque collaboration between state and corporate actors has increased the vulnerability of those who share information on social media platforms. For example: the police use of social media and facial recognition technology to identify and detain protesters at the 2020 Black Lives Matter protests; Uber’s collaboration with Polaris to deputize drivers to report signs of human trafficking; and the world’s largest electronic monitoring company doubling the use of SMARTLink, an app to monitor social media, to surveil people under ICE supervision.
It doesn’t end there. Social media platforms (e.g. Facebook, Twitter, and Snapchat) work with private surveillance companies that collaborate with the state, like Thorn, an organization that supposedly aims to end child sex trafficking. As Kate Zen said on a Hacking//Hustling panel at Harvard, Thorn’s program (Spotlight): “takes escort ads from various different advertising sites and makes it available so that Facebook, Twitter, Snapchat, Pinterest, Imgur… Tinder and OK Cupid all have access to your escort ads. They have access to your faces and your photos if you’ve done any ads.”
What Is Shadowbanning?
Different types of content moderation occurred in early Internet communities. Often, in early Internet communities, the labor of content moderation was undertaken by volunteers who were part of their online communities. These content moderation practices were developed and implemented by the communities they served. These practices and mechanisms of content moderation were “often direct and visible to the user”. These overt moderation actions gave users an opportunity to comply and be in dialogue with moderators.
As platforms grew and began to turn a profit, sex workers, who were some of the first to use these platforms (Patreon and Tumblr, for example), were then deemed high-risk and deplatformed.
Content moderation and shadowbanning are not new; users are just continually learning and reverse engineering how these practices take place on larger platforms that have been rapidly monetized.
Typically, a shadowban means that a user can continue posting as normal, but their posts will be hidden from the rest of the community.” Thus, a shadowban differs from a ban in that a ban is communicated to a user whereas a shadowban is typically not disclosed to the user (and either publicly denied by the platform, or explained away as a glitch or a bug). A shadowban can be understood as a form of platform gaslighting because the platform denies the very real—and felt—practice of shadowbanning users.
A shadowban ensures that content that platforms deem inappropriate, high-risk, or low value speech is invisible to other users, but still allows an account to remain on the platform in order to make ad revenue off of the shadowbanned individual. The shadowban is an integral part of surveillance capitalism, where the user is still on the metadata collection and surveillance matrix. The platform can still surveil, but the user loses their voice.
Despite the invisibilization of high-risk communities on social media platforms, it is much less likely that an individual will be deplatformed. Deplatforming an individual means that the platform is no longer able to generate ad revenue, sell data to data brokerage firms, or provide data to Social Media Intelligence (SOCMINT) companies. Shadowbanning becomes a very powerful tool for platforms to silence dissent while still turning a profit and collaborating with the state to surveil and police communities.
Marginalized communities have been reporting reduced social media visibility for years, but it wasn’t until Trump tweeted about shadowbanning that the phrase entered the mainstream media. A cursory Google search of the term “shadowbanning” will predominantly take you to far right websites complaining about shadowbanning as a tool that specifically targets the far right.
The experiences of sex workers and AOPs that we share in this report confirm that this issue is not a technical glitch, has not been resolved, and is not a conservative conspiracy theory.
Sex workers and AOPs are silenced by platforms at a faster rate than nazis and the alt-right. As Jessie Sage and Juniper Fitzgerald wrote for Tits and Saas, “Apparently, the dudebros of Silicon Valley perceive a naked breast as more ‘sensitive’ than hailing Hitler.”
What We Know About Shadowbanning
What we know about shadowbanning is primarily thanks to community knowledge and the direct experiences of the marginalized communities experiencing harassment, reduced visibility, shadowbanning, and deplatforming. What we know to be true about shadowbanning—and what our research supports—is that shadowbans are felt intensely by the marginalized and radicalized communities that do sex work and movement work. This impact isn’t being discussed outside of our very niche community.
Platforms have rarely, if ever, admitted to shadowbanning. Twitter officially maintains that it does not shadowban. In a response to the publication of this report, a representative from Twitter said:
“Everyone can express themselves on Twitter as long as they don’t break the Twitter Rules. We don’t block, limit, or remove content based on an individual’s views or opinions. In some situations, a Tweet may not be seen by everyone. Per our sensitive media policy, ‘You can share graphic violence and consensually produced adult content within your Tweets, provided that you mark this media as sensitive.’ If people have safe search results on these filters exclude potentially sensitive content. But people have the option to turn it off, or back on, at any time.”
Yet lived experiences from sex workers and AOPs confirm that shadowbanning is happening, despite Twitter’s convoluted denial. Shadowbanning is so prevalent and impactful that community has developed Shadowban Testers, which check Twitter accounts for search suggestion bans, search bans, ghost bans, and reply deboosting. Jessie Sage and Juniper Fitzgerald say that, “shadowban testers are able to determine whether a Twitter user is banned in search suggestions, general searches, and/or in their thread. Essentially, the test detects whether a user’s Twitter handle is suggested to others, whether their handle pops up in a general search, and/or if the user’s entire thread is invisible to other users.”
Despite also denying the practice of shadowbanning, Facebook filed a patent for content moderation practices that could include shadowbanning—or, as they call it, “Moderating content in an online forum”—on July 16th, 2019. The abstract of the patent is vague about implementation and execution, but does seemingly reference shadowbanning:
“For example, the social networking system may receive a list of proscribed content and block comments containing the proscribed content by reducing the distribution of those comments to other viewing users. However, the social networking system may display the blocked content to the commenting user such that the commenting user is not made aware that his or her comment was blocked, thereby providing fewer incentives to the commenting user to spam the page or attempt to circumvent the social networking system filters.”
Other elements of the patent support our research-driven theory of the existence of networked shadowbanning, whereby you may be penalized for the actions of those within your social network.
“[…] identifying users of the social networking system who are connected to the posting user in the social networking system with a specified connection type, wherein the specified connection type is a one-to-one friend connection”.
This patent suggests Facebook may have, or may be considering developing, technology necessary to shadowban entire communities based exclusively on who they know and frequently engage with. This would create an increased chilling effect for already marginalized communities. In our research, we explore how our data suggests the existence of networked shadowbanning practices and how it plays out in sex worker and AOP communities.
Structural Platform Gaslighting, Racist Algorithms & Offline Policing
In 2019, Twitter stated in a blog post, “People are asking us if we shadow ban. We do not.” In this claim, Twitter states that they do not shadowban, which they later contradict in the same blog post: “(although you may have to do more work to find them, like go directly to their profile).” Having to visit someone’s profile to see their tweets, even though you follow them, is an opaque content moderation tactic that decreases the reach of content in ways that are not apparent to the user.
Despite the ubiquitous denial of shadowbanning across platforms, shadowbanning is something that communities vulnerable to policing and surveillance have reported long before these practices gained mainstream recognition. Lynn, a phone sex operator and phone sex coach, told Tits and Sass: “You can try to stay under the radar and not attract the attention of would-be morality police. But, fundamentally, you are at the mercy of unknown policies and algorithm math.” In an investigative report on TikTok’s racist and classist content moderation practices, The Intercept found that fans, “must continue to rely on the Community Guidelines page to guide their conduct, while the actual rules remain always on the verge of revision, revocation — or disavowal via corporate statement.”
Websites like Shadowban.eu exist to help a user tell if they are shadowbanned on Twitter. Shadowban.eu says they conduct shadowban tests by querying the Twitter search for usernames and tweets with at least one response, and analyzing the results for visibility. While there is no way to know the efficacy of tools like Shadowban.eu, many community members rely on it because there is no transparency from the platform.
Despite multiple platforms adamantly denying the practice of shadowbanning, the evidence that these practices are taking place is more than just community knowledge. In 2019, Twitter’s Terms of Service (TOS) changed to reflect what they call “lowered visibility”. On January 1st, 2020, Twitter’s TOS changed again to state that they: “may also remove or refuse to distribute any Content on the Services, limit distribution or visibility of any Content on the service.” In leaked screenshots from a July 14th, 2020 Twitter hack, we see that on the backend, Twitter has accounts tagged as “Trends Blacklist” or “Search Blacklist”. When platforms deny these practices that communities feel, we are left to our own devices to figure out how to stay visible and maintain our platform.
(A screenshot from the backend from July 14th, 2020 Twitter hack.)
When we do not know how the processes that disrupt our ability to make a living and conduct our organizing work, we are left trying to decipher the hidden processes and labor of content moderation and platform policing. Sarah T. Roberts describes this process of obfuscation as “immaterial and ethereal, rather than grounded in the physical word and reliant upon human actors”.
This image is from a digital performance by Veil Machine—a sex worker art collective run by Empress Wu, Niko Flux, Sybil Fury, and friends—to protest social media shadowbanning and the deplatforming of sex workers.
It highlights the sisyphean task of complying to community guidelines that are: intentionally vague and applied differently to different people. This, coupled with the ubiquitous cross-platform denial of shadowbanning, creates an environment of structural platform gaslighting, which mirrors many relationships with power.
This platform gaslighting is continued when accusations of shadowbanning, reduced visibility, or downranking are dismissed as a bug or a glitch. Jessie Sage and Juniper Fitzgerald write, “This practice shames sex workers while insisting that we can be free of any particular platform’s discipline and punishment, so long as we jump through ambiguous hoops designed to ensure our failure.”
Instagram is notorious for sexist, racist, and whorephobic content moderation and platform policing that demote content they find inappropriate or sexually suggestive. When Huffington Post reached out to Instagram, they refused to define the terms and without explanation. At times, Instagram has banned hashtags such as #woman and #curvy. “In this case, #curvy was consistently being used to share content that violates our guidelines around nudity,” said a spokesperson for Instagram. Research has shown that plus-size users’ accounts are flagged for excessive nudity and sexual solicitation more frequently than thin users’ accounts. Also, women of color and queer people experience platform punishment and policing more frequently than their white, cishetero counterparts. In this instance, it is not just sex workers whose content is invisibilized or shadowbanned. With the banning of #curvy and the suppression of any content that goes against white supremecist patriarchal norms, Blackness, fatness, femininity, and feminine sexuality all become a proxy for inappropriate content and whore.
In some cases, individuals following the community guidelines to the best of their ability are still deplatformed and unable to get their accounts back unless they know someone who works at the offending platform. There are social media management companies that charge a large fee in exchange for getting your account back. Sometimes anti-porn vigilantes serve as bad-actors and weaponize a platform’s content moderation system, which leads to the deplatforming of individuals, whether they are complying with the TOS or not. Danielle Blunt of Hacking//Hustling says, “Platforms need to do a better job of keeping marginalized communities safe on their platform. If a troll is able to weaponize your content moderation system to silence the voices of marginalized communities, your content moderation system has failed.” Or perhaps the system is working exactly as intended: deeming communities as high-risk and deserving of removal from public spaces.
In an investigative report on shadowbanning, Paula Akpan writes, “What makes things even more difficult is that almost everything we know about how [social media] functions is guesswork, even for social media experts.” This quote highlights the power imbalance between users and platforms, where blackbox technologies obfuscate the methods used by platforms to inflict violence on, and invisibilize, communities. This practice of platform gaslighting mirrors the way that police routinely deny policing tactics that have been used against Black people and street-based sex workers, such as jump-outs—the same tactics that are now being used on protesters in the 2020 Black Lives Matter uprisings.
In a world where our interactions are increasingly occurring online, it shouldn’t be surprising that content moderation practices and platform punishment can follow us into our offline interactions, too. Similar to platform policing and algorithmic bias, street policing mirrors the biases and racism of its creators. Street policing also relies on the racist and transphobic grouping of high-risk individuals.
In Hacking//Hustling’s Legal Literacy Panel, Lorelei Lee points out that everyone breaks the law and that we mistakenly think that criminalization is about behavior, rather than which communities are being policed. They go on to say, “the way that information gets used against you that isn’t really cognizable in the law, but once they have your information and have you on their radar, they use that information to get more information, to follow you, to trace your contacts, and [this happens] in multiple different contexts.” We see this in predictive policing and risk-assessment technologies that are often deployed without community knowledge, consent, or understanding. These technologies are used to determine what your future behavior might be based on where you live and “perpetuate criminalization through racial and gender bias.”
Even when you are not breaking the law, characteristics about you or your behavior are used as a proxy by law enforcement and platforms. In the U.S., we see this when Black trans women are arrested on sex working charges for occupying public spaces, when condoms are used as evidence of prostitution. In India, we see this with Aadhaar, India’s unique biometric identification programming, where transgender women have limited job opportunities other than sex work, so Aadhaar uses transgender women as a proxy for whore.
With the current global uprisings against the police murder of Black people across the country, we see very concretley how social media can follow us offline—sometimes with deadly consequences. Just recently, there have been reports of protesters getting arrested for organizing or even demonstrating in solidarity with Black Lives Matter. Police and government agents were able to identify these protestors by tracking where they bought their clothing (one protestor in Philadelphia was tracked through Etsy, for example) and running protesters social media livestreams and photos through facial recognition algorithms. These examples show us how censorship and attacks on free speech are facilitated by structural racism (i.e. racist algorithms).
For communities especially vulnerable to policing and state surveillance, Sarah T. Roberts introduces the idea of takedowns and deplatforming as a “negotiated cost” of using social media platforms. For example, Black TikTok content creators report noticing their follower and view counts drop drastically after posting about racism, race, or the Black Lives Matter movement. Some Black content creators report that they have had videos taken down without notice from the platform, leaving them wondering what they did to warrant deletion. Research also indicates that Black Americans who use social media more frequently experience more discrimination than Black Americans who use social media less frequently. This suggests that for Black Americans social media are both sites of oppression and expression.
Surveillance Capitalism and Sex Workers
There is economic value for corporations to weaponize sex workers’ personal data and criminalize their online communications. Like others, sex workers’ personal data is bought and sold, aiding the acceleration of personal data analytics and revenue.
Corporate accumulation of personal data is a “primary mechanism for social manipulation and control in the information age.” In this new era of capitalism “fuelled by personal data,” corporate dataveillance and the personal data industry’s far reach make privacy for all individuals, regardless of whether or not they are AOPs, close to impossible. It is adamant then, to recognize surveillance capitalism as a social justice issue that hinders participation in democratic processes and social life.
Once an algorithm identifies someone as a whore, this label can follow individuals who trade sex (or fit the profile of someone who police think might trade sex) to non-sex work content, cross-platform, and offline. Once an algorithm identifies you as a whore, that stigma and the subsequent experience of shadowbanning and policing, punishes you regardless of your content or employment status. Jessie Sage and Juniper Fitzgerald describe this process: “wildly unsettling that our innovations are co-opted and sold in a marketplace that we’re not even allowed to access. Access to visibility, whether we use it to post pictures of our pets or start the fucking revolution, is indeed a human right. It is a right that is being actively quashed by paternalistic tech companies with help from the surveillance state and moral crusaders.”
This type of punitive content moderation is also applied in the financial technology (fintech) industry. Many sex workers lose access to not only their social media platforms, but their bank accounts and payment processors as well. Sex worker AOP overwhemingly fund their organizing efforts with their sex working income.
One of the largest barriers to mobilizing against surveillance capitalism is the lack of public awareness. However, there has been growing public concern and distrust in the ways corporations handle their users’ personal data. Given a rise in grievances regarding the erosion of privacy, there are several active social justice movements working to illuminate flaws in profit-driven content moderation and surveillance.
While privacy rights and demands for free speech may unify many anti-censorship groups, the fight for sex workers’ rights is too often left out of conversations about digital freedoms. Perhaps it is in moments of political uprising that the impact of content moderation and platform policing become most visible. Amidst a pandemic, our reliance on online spaces to organize is paramount, and the processes by which platforms invisibilize our work and erase our organizing efforts become more visible in the gap they leave behind.
The current anti-privacy policies and related efforts to control communities online will, over time, have an impact that stretches far beyond sex working and AOP communities. These policies will stunt sex technology innovation, prevent LGBTQ youth from accessing resources and community support online, further isolate anyone looking for information on sexual or reproductive health online, and continue to chill speech and disrupt movement organizing.
Sex workers warned about the wide-spread negative impact of FOSTA, but were not heard. We can prevent worsening digital divides in part by listening to sex workers to better understand how they are impacted by platform policing, content moderation, and harmful legislation like FOSTA.
The latest fight against the EARN IT Act shows the importance of recognizing attacks on sex workers’ rights as attacks on our collective rights as a society. For example, many different types of social justice advocates would not be able to safely do their life-changing work without encryption. It is likely that many communities will suffer increased censorship, reduced visibility, and deplatforming if this bill is signed into law.
Eventually, these policies will begin to influence the way everyone—not just sex workers and AOPs—are able to access and use the digital spaces we’ve come to rely on to stay connected, manage our finances, and organize.
It is up to both policy makers and the general public to work in solidarity with sex workers and organizers working towards a free, open Internet for all. As we have said before, an Internet that is safer for sex workers is an Internet that is safer for everyone.
This research explores the intersection of sex workers’ and AOPs’ online experiences, and seeks to better understand how content moderation impacts their right to work and organize. In this research, we ask what shifts are happening in sex workers’ and AOPs’ experiences of content moderation amid civil uprisings, a global public health crisis, and increasing online surveillance. And how does ongoing content moderation impact sex workers’ and AOPs’ ability to work and organize?
In this report, we refer to survey respondents who identified as “activists, organizers and protestors” as AOPs for brevity. We surveyed respondents who identify as only a sex worker, only an AOP, or both a sex worker and an AOP. We allowed respondents to pick which category best fit their lived experience, and allowed respondents to define their experience of being an AOP in their own words from their own perspectives.
We collected data from these three groups in an effort to better understand how sex workers and AOPs experience content moderation—to examine both similarities and differences. We also acknowledge that the work of AOPs is largely underfunded and often supported indirectly with sex work. Within this research, our aim is the centering of sex working AOPs in a broader analysis of social movements and surveillance.
In total there were 262 respondents: 20.99% (55) identified as a sex worker, 32.44% (85) identified as an AOP, 38.93% (102) identified as both a sex worker and an AOP, 7.63% (20) identified as other. In some cases, we combine data from those who identified as a sex worker and those who identified as both a sex worker and an AOP to get data on those who have done sex work (vs. those who have not).
Our sample cannot be generalized to the sex worker or AOP populations overall, but this report does provide a window into how content moderation is currently operating within those populations who use the internet.
We clearly note where our data achieves statistical significance throughout the report. We calculate statistical significance using an industry standard 95% confidence level. Statistical significance means the difference between two groups has less than a 5% probability of simply occurring by chance alone.
We collected data through a web-based survey using SurveyMonkey and distributed survey links online via Twitter and Instagram. We also used our personal connections within the sex worker community to distribute the survey. We chose to use an encrypted data collection service. Our survey did not ask for any personally identifying information beyond whether the respondent is a sex worker, an AOP, or both a sex worker and an AOP. We did not collect respondents’ names, usernames, IP addresses, or any other personally identifying information.
In order to investigate our research questions, we asked respondents about their experiences with social media (Twitter, Instagram, YouTube, etc.), financial technologies (PayPal, Stripe, Venmo, etc.), shadowbanning, account deletion, and deplatforming.
We included a question about COVID-19 to understand how many sex workers have begun doing more online work due to the pandemic and if increased online sex work has impacted respondents’ experiences with shadowbanning. We also included questions about respondents’ engagement with Black Lives Matter and community organizing.
We felt that it was important to use a mixed methods research model for data collection so that we could compare qualitative and quantitative data, ensuring our interpretations reflected the experiences of the communities we are studying. By offering some open-ended questions we were able to collect valuable feedback in respondents’ own words about their relationship to online surveillance and content moderation.
We analyzed quantitative data in six different ways: all aggregate data, all data by segment (sex worker vs. AOP vs. both), all data by sex workers who have started doing more online work due to COVID-19, all data by sex workers posting about Black Lives Matter, all data by people who report having been shadowbanned, and all data by those who have done sex work.
Our findings suggest that being a sex worker online and engaging in activism online act as compounding factors, with sex working AOPs disproportionately impacted by the negative effects of content moderation and surveillance capitalism. This means that platforms police, censor, and deplatform AOPs who support their organizing work through sex work more harshly.
Respondents consisted of online users with some degree of digital literacy, and access to social media and the Internet. While their feedback provides vital information about the impacts of content moderation on sex working communities, it does not fully capture the extent to which new surveillance technologies are targeting sex workers online and offline.
Regarding the compounding effect of being both a sex worker and an AOP as opposed to being only a sex worker or only an AOP: this implies to us that if we were to conduct this research with a broader civilian population, these numbers would be even more dramatic. As we discuss in our Limitations section, this research could benefit from a civilian comparison group.
1. We consistently found that people who identified as both a sex worker and an AOP experienced the negative impacts of platform policing both more intensely and more frequently. Sex workers are significantly more likely (30.77%) to report they have been shadowbanned on social media while AOPs are significantly less likely to report the same (12.82%). Of those who identify as both a sex worker and an AOP, an incredible 51.28% report they have been shadowbanned.
2. Sex workers who started doing more online work due to COVID-19 experienced significantly more punitive platform policing than other respondents. 71.14% of people who have done sex work have started doing more online work due to COVID-19. We found that nearly every form of shadowbanning and deplatforming we asked about was more prevalent among sex workers who had started doing more sex work online due to in-person COVID-19 restrictions—even more so than sex workers who already did sex work online.
3. Sex workers who shared original tweets about Black Lives Matter from an account where they also post about sex work were significantly more likely to suffer platform policing. They say they have: noticed a difference in the visibility of their content, posts, or profile since the end of May 2020 (44.30%); and lost access to a financial technology (e.g. PayPal, Venmo, Square Cash, etc.) (51.90%).
4. Respondents who identified as both a sex worker and an AOP demonstrated the most chilled speech. 82.5% said they have avoided posting content for fear of being kicked off, shadowbanned, or facing legal action. Only 44.19% of AOPs have avoided posting content for fear of being kicked off, shadowbanned, or facing legal action, which is significantly less than their sex working peers (68.75%).
5. Movement work is restricted most severely for those who are both a sex worker and an AOP. A sentiment analysis of our qualitative data shows a compounding effect where sex workers who also identified as AOPs experienced the most severe forms of platform punishment. Sex workers and sex working AOPs consistently described severe levels of paranoia and chilling effects that non-sex working AOPs did not.
6. Losing access to financial technologies reduces an individual’s ability to earn a living, and disrupts movement work and mutual aid efforts. 66.13% of sex workers who had been deplatformed from a financial technology reported that it impacted their ability to do sex work. 36.67% of respondents who reported being deplatformed from a financial technology reported that it impacted their ability to do movement work or community organizing.
7. Sex workers are experiencing catfishing and content theft at alarmingly high rates. 43.75% of sex workers and 46.43% of sex working AOPs report having had their images or content used for a fake account that they did not run or provided false information (e.g. having a catfishing account steal their photos). AOPs who do not do sex work are significantly less likely to say the same (14.29%).
8. Sex workers are barred from accessing the marketing tools non-sex working communities use to build their small businesses. Whether you’re looking at social media platforms, website hosting, or financial technologies, the world of small business limits—if not excludes—sex workers at every stage. Sex workers continue to build thriving small businesses despite this barrage of marketing barriers. With so many sex workers funding their movement work through sex work, this has a compounding negative effect on community.
1. Sex Workers vs. Activists, Organizers, or Protesters (AOPs) vs. Both
We consistently found that people who identified as both a sex worker and an AOP experienced the negative impacts of platform policing both more intensely and more frequently.
Compared to only sex workers, sex working AOPs are significantly more likely to say they: use more than one social media account (93.48%); noticed a difference in the visibility of their content, posts, or profiles since the end of May 2020 (41.86%); have ever had an issue using financial technologies to community organize or share money with community (30.12%).
Compared to only sex workers and only AOPs, sex working AOPs are significantly more likely to say they have: received a warning that their social media account is close to deletion (48.91%); lost access to a financial technology (e.g. PayPal, Venmo, Square Cash, etc.) (50%).
Compared to only AOPs, sex working AOPs are significantly more likely to say they have been deplatformed or kicked off of a social media account (45.56%).
While AOPs still report feeling the negative and disruptive impacts of platform policing, it played a significantly lower role in their experiences of navigating online spaces vs. their sex working peers.
Compared to sex workers and participants who identified as both sex workers and AOPs, AOPs were significantly less likely to say they: are familiar with the term shadowbanning (87.67%); have ever been shadowbanned (34.88%); have ever had a post removed from social media (46%); have ever found that their username does not show up in searches (i.e. “name suggestion banned”)(22%); have ever had their images or content used for a fake account that they did not run or provided false information (e.g. having a catfishing account steal photos) (14.29%); have avoided using specific words to avoid being shadowbanned (23.81%); have ever avoided posting content for fear of being kicked off, shadowbanned, or facing legal action (44.19%); have ever had content that does not violate a sensitive media policy marked as sensitive media on their profile (41.86%); and have ever had their profile set to 18+, 21+, or NSFW (13.95%).
The compounding effect of being both a sex worker and an AOP is repeated throughout the report. However, those who have done sex work are far and away the most severely impacted segment.
97.95% of those who have done sex work are familiar with the term shadowbanning while only 87.67% of those who have not done sex work are familiar with the term. 69.57% of those who have done sex work report they have been shadowbanned while only 34.88% of those who have not done sex work said the same.
63.31% of those who have done sex work report their content has been repressed in the timeline while only 54.90% of those who have not done sex work said the same. 67.14% of those who have done sex work have had a post removed from social media while only 46% of those who have not done sex work said the same. 41.84% of those who have done sex work said they have received a warning that their social media account is close to deletion while only 17.65% of those who have not done sex work said the same. 58.57% of those who have done sex work said they have found their username does not show up in searches while only 22% of those who have not done sex work said the same. 72.86% of those who have done sex work said they have experienced social media platforms suppressing their friends’ content from their timelines while only 62.75% of those who have not done sex work said the same.
41.01% of those who have done sex work said they have been deplatformed or kicked off of a social media account while only 21.57% of those who have not done sex work said the same. 45.45% of those who have not done sex work were able to get their accounts back after being deplatformed from social media while only 7.27% of of those who have done sex work said the same. 81.82% of those who have not done sex work made another account after they were deplatformed on social media while only 69.09% of those who have done sex work said the same.
45.45% of those who have done sex work have had their images or content used for a fake account that they did not run or provided false information (e.g. catfishing) while only 14.29% of those who have not done sex work said the same. Those who have done sex work and those who have not had equal success in having the catfishing account removed (40.98% and 40%, respectively). However, 33.33% of those who have done sex work said when their images were used for catfishing, they were simultaneously shadowbanned so the fake account showed up first while 0% of those who have not done sex work said the same.
54.76% of those who have done sex work avoid specific words to avoid being shadowbanned while only 23.81% of those who have not done sex work said the same. Similarly, 77.34% of those who have done sex work avoid posting content for fear of being kicked off, shadowbanned, or facing legal action while only 44.19% of those who have not done sex work said the same. This indicates that the ways that platforms police sex work creates a greater chilling effect in sex worker communities.
67.69% of those who have done sex work have had content that does not violate a sensitive media policy marked as sensitive media on their profile while only 41.86% of those who have not done sex work said the same.
74.19% of those who have done sex work are noticing trends in the suppression of information on social media while only 66.67% of those who have not done sex work said the same.
Having experience trading sex significantly increases the punitive impact of platform policing. In many instances, having done sex work doubled the impact of punitive content moderation practices when compared with those who exclusively identified as AOPs. It is likely that this phenomenon would be even greater if compared with a sample of the general population, who are neither sex workers nor AOPs.
While no social media platform currently outright admits to shadowbanning, sex workers and AOPs are acutely aware of how this practice impacts their lives and disrupts their movements. One sex worker respondent said about shadowbanning, “I think it’s when an algorithm prevents your content from actually being seen, but you’re not told. I don’t know much about how it works.” The fact that users don’t know much about the process of shadowbanning is by design, and this sex worker’s understanding is still more than platforms will admit. Shadowbanning is also sometimes referred to as ghost banning, algorithmic curation, and reduced visibility.
Sex workers and AOPs are working to unpack the outputs of the blackbox of shadowbanning and algorithmic curation while platforms are simultaneously denying using these mechanisms to moderate content.
Platform circumlocution of shadowbanning:
- “We believe strongly in being impartial, and we strive to enforce our rules impartially. We do not shadowban anyone based on political ideology. In fact, from a simple business perspective and to serve the public conversation, Twitter is incentivized to keep all voices on the platform.” -Jack Dorsey, testimony to U.S. Congress, September 2018.
- “This isn’t something that we do,” said A.B. Obi-Okoye, a spokesperson for TikTok, in an email to Digital Trends. “We make our Community Guidelines publicly available and encourage users to review them so they can understand what type of content isn’t permitted on the platform. Users also have the ability to appeal decisions they believe to have been made in error.”
- In June 2020, Adam Mosseri, Head of Instagram, announced in a blog post that the company would be taking a harder look at “how our product impacts communities differently” with a focus on harassment, account verification, distribution, and algorithmic bias. “We need to be clearer about how decisions are made when it comes to how people’s posts get distributed. Over the years we’ve heard these concerns sometimes described across social media as ‘shadow banning’ – filtering people without transparency, and limiting their reach as a result. Soon we’ll be releasing more information about the types of content we avoid recommending on Explore and other places.”
As seen in the quotes below, sex workers and AOPs have deep knowledge about how shadowbanning and content moderation processes work, and how they impact their lives.
Respondents were asked if they knew what shadowbanning was and how they would define it. Only respondents who said they were familiar with the term were shown this question. We allowed respondents to define this experience for themselves, and the results below reflect that.
Respondents’ definitions of shadowbanning:
“Shadowbanning prevents you from showing up in suggestion searches and also limits your online presence within your targeted audience. It affects the amount of interaction with potential clients and limits the social media presence of your business or brand.” -Sex worker respondent
“All I know is that when someone’s account is shadowbanned then they are basically invisible to anyone who does not follow them or have some other direct link. The person who is shadowbanned gets no notification or indication that this has happened.” -Sex worker respondent
“Forcibly reducing the visibility of someone. A roundabout way of suppressing the voice of particular people, without being so blatant as to delete their account.” -Sex worker respondent
“Being removed from algorithms/search results. Not having comments shown. Not showing up in tagging. Some combination of that. Being made less visible without being told that’s the case.” -Sex worker respondent
“Shadowbanning is the opaque practice of social media platforms that ‘reduces’ visibility of content the platform deems ‘high-risk’. This can result in racist, fatphobic, misogynistic, ableist and transphobic ways. And can also be [activated] due to malicious user reporting.” -Sex worker and AOP respondent
“It is a process where people who are seen as ‘problematic’ to social media corps will be de-prioritized by the algorithm. Accounts won’t show up in general searches unless you have the specific @, and accounts will have posts stop appearing in people’s timelines. The biggest one has also been the banning of specific hashtags, which prevents workers from organising and [allows] for non-linear communication to occur.” -Sex worker and AOP respondent
“I have been shadowbanned on both Twitter & IG—my engagement is shit when it happens… On IG, I had my old account of 10 years deleted without warning. It wasn’t all sex work stuff, it was a decade worth of personal memories/a digital photo album I will never get back. Now my new account I never use hashtags—til I used an #ripgeorgefloyd hashtag & my account got shadowbanned for that for some reason??? Ridiculous & racist.” -Sex worker and AOP respondent
“You aren’t banned from the platform, but people cannot search you and analytics from the platform do not work in your favour, so that people are unable to find you easily. It’s like Internet purgatory.” -Sex worker and AOP respondent
“When a social media service, through automated or manual systems, reduces the visibility of your original content – often to avoid offending moderates who don’t want to stumble onto content they are bigoted against.” -Sex worker and AOP respondent
“Being allowed to stay on a platform but having all your activity hidden from view.” -AOP respondent
Types of Shadowbanning
Despite the clandestine practice of shadowbanning, social media users have developed their own criteria for identifying when shadowbanning occurs and in which ways. Based on the data collected in this survey, the following are the most common occurrences of shadowbanning.
Search Suggestion Ban: Respondents describe their social media usernames not coming up in the search bar until the entire username has been typed out or disappearing entirely for people who do not already follow them. For non-shadowbanned accounts, the feature would auto-fill the likely username within a few characters. This can also occur with hashtags.
Decrease in Follower Engagement: Respondents described a sudden decrease in likes, comments, and story views disproportionate to their follower size and typical engagement metrics. This can occasionally happen incrementally over a period of weeks, but is more often sudden. Often, the posts of those who reported being shadowbanned do not show up in the feeds of their followers at all, and are seemingly deprioritized from the algorithm altogether.
Feature Block: Some respondents describe not being able to use certain features that allow them to engage with other users. Some report this as an inability to like or reply to the posts of other users, or the inability to have your own social media username linked to a post.
Temporary Bans: For some respondents, shadowbanning is temporary. Respondents described that their sudden decrease in follower engagement or lack of access to certain features lasted a few days or a few weeks before suddenly returning to normal. For some, altering how they post (strictly following TOS or changing whatever behavior preceded the ban) was credited for the ban being lifted. All of this occurs without notification or explanation from the platform.
Other reported symptoms of shadowbanning include: a sudden decrease in follower count, the involuntary unfollowing of another shadowbanned account, having posts deleted by a platform without notification, and the inability to post at all. Some users reported using sites that will track your analytics to verify if your account is shadowbanned. Others reported verifying with peers that they’re search suggestion banned or that their posts are not showing up in the feeds of their followers.
“Feels like I’m posting to the void. Used to get 100s of likes. Sometimes struggled to get 10.” -Sex worker respondent
“I was completely unable to follow, unfollow, like, or make any posts on my Instagram for a time, for seemingly no reason.” -Queer activist respondent
“I have a search suggestion ban for ages, which means my account won’t show up in search results. I’ve been completely shadowbanned a few times and my content wouldn’t show up in the timeline and people wouldn’t even get notifications from my replies.” -Sex worker and AOP respondent
Sex workers are significantly more likely (30.77%) to report they have been shadowbanned on social media while AOPs are significantly less likely to report the same (12.82%). Of those who identify as both a sex worker and an AOP, an incredible 51.28% report they have been shadowbanned.
Our data suggests that automated marketing features may allow platforms to identify bots, making individuals who use them more likely to be shadowbanned. Sex workers who have linked their social media to a sex work advertisement are significantly more likely to report they have been shadowbanned (88.42%). Sex workers and AOPs who do not automate or schedule posts from their primary social media account are significantly less likely (83.78%) to report they have been shadowbanned.
Sex workers and AOPs who report they have been shadowbanned on social media are significantly more likely to say they: believe their content has been repressed in the timeline (e.g. posts not showing up visibly on a newsfeed) (79.49%); have had a post removed from social media (68.10%); have received a warning that their social media account is close to deletion (43.59%); have been “search suggestion banned” (73.28%); have had difficulty finding or not been able to find a friend, comrade, or sex worker in social media searches (e.g. their name not showing up) (82.76%); have experienced social media platforms removing individual followers without their consent (62.07%); have been deplatformed or kicked off of a social media account (44.83%); are noticing trends in the suppression of information on social media (80.95%); have had their images or content used for a fake account that they did not run or that provided false information (e.g. a catfishing account) (45.13%); and have lost access to a financial technology (e.g. PayPal, Venmo, Square Cash, etc.) (41.23%).
Sex workers and AOPs who report they have been shadowbanned on social media are significantly more likely to report they have experienced social media platforms suppressing their friends’ content from their timelines (81.90%).
42.48% of sex workers and AOPs who report they have been shadowbanned on social media have noticed a difference in the visibility of their content or profile since the end of May 2020. This is significantly higher than sex workers and AOPs who do not report they have been shadowbanned.
Of the total sample, 34.18% reported a difference in the visibility of their content, posts, or profile since the end of May 2020. While many who identify soley as sex workers report frustration leading to a decrease in their use of social media sites, those engaged with activism in any form say that they still rely on social media for news and to amplify their movement goals. Those who identify as both sex workers and AOPs report that they’ve scaled back their posts for marketing, prefering to focus on movement work around Black Lives Matter and COVID-19 mutual aid.
“I can’t care about boners right now so it’s not my concern. It’s to help get out info and less about marketing. It’s to connect for the cause.” -Sex worker and AOP respondent
“I feel more connected to it for political purposes. I’m having a hard time balancing activism and professional content, but to stand by and be complicit in white silence is not an option for me.” -Sex worker and AOP respondent
Sex workers and AOPs who report they have been shadowbanned on social media are significantly more likely to say their profile is set to be 18+, 21+, or NSFW (60.55%). This may suggest that self-identifying as an 18+, 21+, or NSFW account mimics the impact of shadowbanning or increases the risk of shadowbanning.
Sex workers and AOPs who report they have been shadowbanned on social media are significantly more likely to say: they have had content that does not violate a sensitive media policy marked as sensitive media on their profile (71.82%). This may suggest that sensitive media flags are a form of shadowbanning.
Sex workers and AOPs who report they have been shadowbanned on social media are significantly more likely to change their behavior to avoid further penalization. They are significantly more likely to say there are specific words that they avoid using to avoid shadowbanning (56.60%). Similarly, sex workers and AOPs who report they have been shadowbanned on social media are significantly more likely to say they have avoided posting content for fear of being kicked off, shadowbanned, or facing legal action (77.06%).
Sex Work and COVID-19
The rapid onset of worldwide social distancing and public health measures in response to COVID-19 has pushed the world towards an even heavier reliance on digital technologies. Simultaneously, measures of surveillance and censorship through those same digital technologies—which are now considered vital, both for personal use and commerce—have been expanding.
This new reality is pushing sex workers further away from necessary, life-saving resources, such as community and capital, and into more visible forms of sex work with fewer security measures. 71.14% of people who have done sex work have started doing more online work due to COVID-19.
The pandemic has been particularly devastating for people who sell sex as the sex industry has been hit hard by censorship and shutdowns. As we learned in our previous research, Erased, many sex workers have no social or financial safety nets to fall back on. COVID-19 has highlighted sex workers’ long-standing, vibrant mutual aid networks and abilities to support one another. But it has also exposed the lack of material support in place for people in the sex trade.
Our research suggests that the impact of using social media for sex work extends cross-platform, both between social media platforms as well as between financal technologies. This violence can also transcend digital spaces when it falls into the hands of state-based actors, or exacerbates financial inequities and isolation. With COVID-19 and our new digitally-dominant reality, this causes further violence to marginalized communities.
We found that almost every form of shadowbanning and deplatforming we asked about was more prevalent among sex workers who had started doing more sex work online due to in-person COVID-19 restrictions—even more so than sex workers who already did sex work online. These results indicate that the more active and visible a sex worker is on a platform, the more likely they are to be shadowbanned.
Sex workers who have started doing more online work due to COVID-19 are significantly more likely to report they: use social media for sex work (96.26%); have more than one social media account (93.07%); have linked their social media to a sex work advertisement (82.24%); have experienced social media platforms suppressing their friends’ content from their timelines (78%); have had content that does not violate a sensitive media policy marked as sensitive media on their profile (76.34%); have had a post removed from social media (75%); have experienced social media platforms unfollowing people without consent (58%); have been deplatformed or kicked off of a social media account (48%); have received a warning that their social media account is close to deletion (47.52%); and have experienced social media platforms unliking posts without consent (39%).
But sex workers aren’t just seeing the impact on social media platforms. The effects of increased sex work online span cross-platform to financial technologies as well. Sex workers who have started doing more online work due to COVID-19 are significantly more likely to report they have: had a payment blocked, cancelled, or flagged because of sex work (62.62%); lost access to a financial technology (e.g. PayPal, Venmo, Square Cash, etc.) (44.79%); had an issue using financial technologies to community organize or share money with community (29.47%).
These patterns among sex workers who have started working online more due to COVID-19 are confirmed by sex workers who are not currently investing more heavily in online work. Sex workers who have not started doing more online work due to COVID-19 are significantly more likely to say: they have not been name suggestion banned (38.46%); they are not sure if they have had content that does not violate a sensitive media policy marked as sensitive media on their profile (30.56%); they do not believe their content has ever been repressed in the timeline (e.g. posts not showing up visibly on a newsfeed) (20.51%).
For some sex workers, “more online work” means “online work for the first time”. One sex worker who started doing more online work because of COVID-19 describes their relationship to moving to online work as:
“I haven’t really had any relationship to social media as a sex worker yet (I was a stripper before covid, and very offline in that persona), so now that I might be moving online, I’m just trying to figure out how to navigate it safely. I guess I’d describe my feelings about it as totally freaked out and out of my depth.” -Sex worker respondent
The learning curve for sex workers completely new to online work is steep as they attempt to navigate the digital security concerns that come with highly visible sex work and content moderation guidelines that are neither public nor uniformly enforced. They’re learning—exclusively through trial and error, and shared community knowledge—how to maintain the level of digital hygiene required to use the mainstream social media platforms and financial technologies that other businesses are built upon.
There is a direct correlation between sex workers increasing online work and losing access (partially or completely) to both social media platforms and financial technologies. Sex workers are witnessing this suppression of their peers’ content, but remain less sure about the suppression of their own content—operating in a fog of platform gaslighting that only compounds when you layer activism on top of sex work.
Sex Work and Black Lives Matter
After the start of the 2020 Black Lives Matter uprisings, Instagram users found themselves unable to post about #BLM. Users reported seeing “action blocked” messaging without context when attempting to post about ongoing protests and related human rights issues.
On June 1, 2020, Instagram’s Comms team tweeted: “We have technology that detects rapidly increasing activity on Instagram to help combat spam. Given the increase in content shared to #blacklivesmatter, this technology is incorrectly coming into effect. We are resolving this issue as quickly as we can.” This public acknowledgement from Instagram highlights what countless sex workers, AOPs and other users have alleged for years: it’s possible for platforms to suppress content from certain communities, limiting the reach of their collective voice. This is just one example from one social media platform of how automated processes suppress content. Our research shows that sex workers and AOPs face this type of censorship cross-platform year-round.
Our research shows that sex workers who were sharing Black Lives Matter and protest-related content on their sex worker social media accounts were significantly more likely to experience punitive content moderation measures, including shadowbanning. This group was less reserved when talking about sex work and activism, which suggests a compounding effect of platform policing for sex workers and AOPs.
Sex workers who also identified as AOPs were significantly more likely to say they have shared things about Black Lives Matter from an account where they also post about sex work in the form of original tweets (vs. just retweets) (76.47%). Sex workers are significantly more likely to say they have only shared things about Black Lives Matter from an account where they also post about sex work in the form of retweets (vs. original tweets) (52.63%).
“I have more or less stopped using social media to promote myself and have transitioned to using it to promote BLM, anti-ICE, and other social justice/human rights information. I mostly retweet things at this point.” -Sex worker and AOP respondent
Sex workers who shared original tweets about Black Lives Matter from an account where they also post about sex work were significantly more likely to say they have: linked their social media to a sex work advertisement (94.05%); and tried running paid search or social ads (26.58%). This implies that this group may be more active and visible with their online sex work, sharing more metadata and content that may facilitate an increase of cross-platform punishment and networked shadowbanning.
Sex workers who shared original tweets about Black Lives Matter from an account where they also post about sex work were significantly more likely to say they have: avoided using specific words to avoid being shadowbanned (63.51%); experienced social media platforms unliking posts without their consent (41.46%); and noticed a difference in the visibility of their content, posts, or profile since the end of May 2020 (44.30%).
“I said ‘fuck it, nobody is seeing my content anyway’ and started posting less + started posting less sex work-specific things and more political things that I care about (like BLM-related content).” -Sex worker and AOP respondent
As we’ve seen throughout our research, the impact of this activity extends cross-platform from social media to financial technologies (and vice versa). Sex workers who shared original tweets about Black Lives Matter from an account where they also post about sex work were significantly more likely to say they have: had a payment blocked, cancelled, or flagged because of sex work (63.53%); lost access to a financial technology (e.g. PayPal, Venmo, Square Cash, etc.) (51.90%); and had an issue using financial technologies to community organize or share money with community (30.26%).
Again, we see evidence of a potential networked web of cross-platform scraping, sharing of data, and shadowbanning. Our research suggest that platforms may use algorithms and automated processes to share identity unifying information with one another, facilitating the cross-platform deplatforming of sex workers and AOPs.
38.9% of respondents identified as both a sex worker and an AOP—easily our largest segment. Many AOPs fund their organizing through sex work. When platforms police, censor, and deplatform both sex workers and AOPs more harshly, this creates a compounding effect.
This data suggests a mutually conflicting dilemma for sex worker and AOP visibility. The more visible you are as a sex worker and/or an AOP, the more platform policing you will feel and the more your work will be suppressed. Sex workers and AOPs see shadowbanning, which platforms often publicly call an “isolated glitch”, through the glass-ceiling of content moderation.
Chilled speech is when an individual’s speech or conduct is suppressed by fear of penalization at the interests of a party in power (e.g. the state, a social media platform, etc.) It can ultimately limit free speech and open discourse. As we saw with FOSTA-SESTA, content moderation practices chill speech through fear, whether or not punishment actually ever occurs. When sex workers and AOPs experience chilled speech, we see that movement and organizing work suffers. Chilled speech reduces individual and community responses to violence and injustices due to fear of punishment.
44.19% of AOPs have avoided posting content for fear of being kicked off, shadowbanned, or facing legal action, which is significantly less than their sex working peers (68.75% of sex workers said the same). Respondents who identified as both sex workers and AOPs demonstrated the most chilled speech with 82.5% saying they have avoided posting content for fear of being kicked off, shadowbanned, or facing legal action.
Similarly, 23.81% of AOPs said there are specific words that they avoid using to avoid being shadowbanned, which is significantly less than their exclusively sex working peers (54.35%). Again, respondents who identified as both sex workers and AOPs demonstrated the most chilled speech with 55% saying there are specific words that they avoid using to avoid being shadowbanned.
“My speech is chilled. I am unable to say what I would like to say and resort to using steganography or weird spellings. It makes me hesitant before posting. Even when advocating for sex worker rights, I don’t spell out ‘sex worker’ because it is an instant shadowban, I’ve noticed.” -Sex worker and AOP respondent
45.45% of AOPs were able to get their accounts back after they were deplatformed, but only 6.25% of sex workers and 7.69% of people who identified as both were able to get their accounts back. If you haven’t done sex work, you are over 5x more likely to get your account back after being deplatformed.
81.82% of AOPs made a new account after being shadowbanned while only 62.50% of sex workers did the same.
We see (warranted) themes of hypervigilience among those who trade sex as fears of deplatforming dictate what sex workers post on social media every single day. This further chills speech for sex workers, disabling their use of social media as a marketing channel.
“On Instagram I have to be very careful about the content I post. There are so many banned hashtag words. And I have to be careful not to show too much skin. I also cannot directly link any of my NSFW content to my Instagram account. I try to use Twitter for that, but I don’t currently have a big enough following there to rely only on that platform.” -Sex worker respondent
Even non-sex work content is being chilled—often more so if posted by a sex worker—implying that platforms are policing people, not content. One sex worker said they avoid talking about rape and bad date lists for fear of being deplatformed, shadowbanned, or facing legal action. Another said:
“Content moderation often entails shadowbanning/intentional algorithmically reduced engagement, issuing policy violations regardless of actual post content, banning content promotion for sex workers and NSFW art, as well as citing violations for language used in posts, even if such language is being used in opposition to users threatening their safety. Context appears to matter less than content. Anything that offends the idea of a family friendly image or outing white supremacists is typically treated with harsher punishment than users who openly issue death and rape threats or violent ideologies & actions. Admins of Twitter and Instagram give their tacit approval to certain violent groups while suppressing harmless users, such as sex workers and activists.” -Sex worker and AOP respondent
For fear of doxxing, shadowbanning, deplatforming, and other attacks from racists, homophobes, and fascists, sex workers and AOPs report avoiding:
- “Nudes and politics”;
- “Text [from] and links to scientific studies”;
- “Anything related to the LGBT community”;
- “Defending my rights as a transwoman”; and
- Explicit content and BDSM/kink content (including links to owned websites or OnlyFans).
Chilled speech has had such an overwhelming impact on community that another sex worker reports:
“I have avoided posting at all tbh.” -Sex worker respondent
Organizing Under a Shadowban & Disruption of Movement Spaces
As power becomes increasingly concentrated in the hands of Big Tech, collective action is further entangled in the web of mass datamining, chilled speech, and surveillance capitalism. AOPs from communities vulnerable to state and platform policing disproportionately face the paradox of having to mobilize via new technologies that put them at risk, while simultaneously making them hypervisible.
In the midst of a pandemic—during which many sex workers are relying on digital technologies not only to make an income and survive, but to collectively organize as well—the rise in hypervisibility of digital sex worker organizing is a double-edged sword. Media attention directed at mutual aid and relief funds for sex workers, though often well-intentioned, has further exposed these individuals and movements, making them even more vulnerable to digital repression and harsher criminalization.
Especially under these increasingly harsh conditions, we believe it was important to include an open-ended question asking respondents to describe their overall relationship to surveillance and social media. In this question, we received 134 responses: 33 sex workers, 35 AOPs, 66 both. A sentiment analysis of this qualitative data shows a compounding effect: movement work is restricted most severely for those who are both a sex worker and an AOP.
Though all three groups of respondents reported distrust in Big Tech and surveillance, sex workers and sex working AOPs continually described severe levels of paranoia and speech chilling that non-sex working AOPs did not. One sex worker respondent referred to social media as a “minefield” while another commented: “I never feel free.” Throughout the data, sex working respondents detailed the ways that surveillance disrupts their movement work:
- “As a sex worker, I constantly have to censor my posts or choose to risk a shadowban for a post that I want to share. My relationship could be described as ‘frustrating as hell’. Social media is necessary as a business model, but also as a way to connect those other sex workers and activists. Having platforms strategically surveil or suppress our posts is purposefully reducing our visibility, our needs from society, and calls to action.” -Sex worker respondent
- “It terrifies me that my inherent right to privacy is constantly thwarted and that organizing efforts are as well.” -Sex worker respondent
- “Twisted. You need it to keep up and stay in touch and organize. But at the same time it is very much the system you are fighting.” -Sex worker respondent
Both sex workers and AOPs largely referred to their use of digital tools as a “love-hate” relationship. Several respondents described social media as something that would be a useful tool if they were able to use it how they wanted and needed to. As Zeynep Tufekci illustrates in Twitter and Tear Gas, there is both power and fragility in the use of social media for political mobilization; in the case of AOPs, especially sex working AOPs, social media is often a “necessary evil”. One sex working AOP detailed their relationship to social media and surveillance as “powerful and paranoid”, which parallels how Tufekci frames the power and fragility of networked protest.
What is missing from Tufecki’s analysis, but captured by our data, is the compounding effect highlighted by sex working AOPs: the intersection of identities between sex worker and AOP operates under an even harsher punitive ecosystem, ruled by capitalist systems of data surveillance and regulation.
“As a sex worker and activist, my relationship to both is complicated since social media offers a platform with which to promote my work and grow my business, as well as connect with my community as an activist, but can also be used against me since my identity as a sex worker/activist places me at greater risk for surveillance and deplatforming.” -Sex worker and AOP respondent
Another sex working AOP describes their frustration organizing and working in a whorephobic online ecosystem. They say, “I’d love to be able to post organizing info & sex work info & be able to hashtag it in an appropriate manner for the people who need to see it to be able to find it effectively—without it resulting in censorship.” Another respondent shares, “I’m very wary of (social media), but I also feel somewhat powerless about it. How can we collectively fight these algorithms of oppression when we rely on them to organize in the first place?”
Despite the compounding effect of being both a sex worker and an AOP, AOPs that do not do sex work still often detailed a similar hypervigilance around using digital technologies. This suggests that criminalization policies and measures of surveillance targeted at people who sell sex may have a ripple chilling effect, setting a standard for how other marginalized and politically active communities can be similarly monitored.
While social media platforms are frequently seen as providing more access to movement participation, our data makes us ask the question: are these platforms really accessible? And if so, for whom?
The exploitation of sex worker precarity in the digital age makes movement work incredibly challenging. Many sex workers fund their community work with sex work. When sex workers’ ability to make income is halted, their movement efforts are similarly compromised.
In this current political moment, even sex workers’ access to elected representatives is reduced by online platforms. For example, town halls are being hosted on Facebook, which sex workers are increasingly barred from. This limits the ability for sex workers, especially sex working AOPs, to participate in democratic society.
Disruption of the flow of capital interferes with an individual’s ability to make a living with online sex work, accept online payment for in-person sex work, and coordinate mutual aid efforts. Both financial and digital literacy barriers block sex workers, disproportionately sex workers already on the margins, from the digital tools that remain intact or manage to pop up, many of which are higher-cost and highly surveilled. In our previous report, Erased, 33% of online sex working respondents reported losing access to a payment processor (such as PayPal, Venmo, or Square Cash) and 78% of street-based respondents reported not having access to a bank account.
“It makes me extremely wary and prepared to be deplatformed. I diversify my financial tools, and try not to keep ‘too much’ money in any given financial tool. Diversify my bank accounts.” -Sex worker and AOP respondent
In this report, we found that 32% of all respondents have lost access to financial technologies (such as PayPal, Venmo, Square Cash, etc.) Respondents who have reported losing access to a financial technology reported PayPal (39.66%), Square Cash (18.97%), and Venmo (6.9%) as the most frequent platforms they’ve lost access to.
“When I lost access to CashApp, I had to take deposits via much riskier (more identifiable) methods instead. I occasionally used other payment processors instead, resulting in 30-45% lost income due to their high [processing] fees.” -Sex worker and AOP respondent
66.13% of sex workers who had been deplatformed from a financial technology reported that it impacted their ability to do sex work. 36.67% of respondents who reported being deplatformed from a financial technology reported that it impacted their ability to do movement work or community organizing.
“Not having a reliable place to hold funds and receive electronic payments is a huge issue and barrier to mutual aid. It also forces the money to sit directly in someone’s account, which can mess with taxes because it looks like income.” -Sex worker and AOP respondent
Losing access to financial technologies disrupts the flow of money within community and one’s ability to organize, but seems to be more severe for sex workers. 40.74% of sex workers have lost access to a financial technology while only 14.29% of those who have not done sex work said the same. 55.33% of people who have done sex work have had a payment blocked, cancelled, or flagged because of sex work.
38.46% of those who have done sex work said that being deplatformed from financial technology impacted their movement work or community organizing while only 28.57% of those who have not done sex work said the same. 77.36% of those who have done sex work said that being deplatformed from financial technology impacted their sex work.
“It is impossible to share money with community because no one has access to the same accounts anymore. Also, I am constantly sending over the weekly limit when I am redistributing money.” – Sex worker and AOP respondent
Cash is king, but not amid a pandemic, which is forcing sex workers and AOPs to work and organize in digital spaces. As we increasingly come to rely on financial technologies, we need to be discussing who has access to these technologies and who is being systematically removed from them.
Catfishing is the act of creating an online persona by posting photos or videos of another person, with the intention of interacting with others while pretending to be the person pictured. Popularized by the documentary and TV show by the same name, the term was coined specifically to describe the act of manipulating another into believing the catfish is someone who they’re not.
In this research, we are examining the effects of being the person whose photos have been stolen by a digital impersonator, primarily for the purposes of making money by non-consensually using the images of a sex worker.
Among the more insidious aspects of online security and safety, digital impersonation is a problem that social media platforms don’t seem to prioritize as heavily. 27.69% of all respondents reported having their photos used by digital impersonators.
Those who identified as sex workers were more likely to have their photos stolen and used on catfishing profiles. Many attributed this to impersonators creating sex working profiles or advertisements—using stolen photos—to solicit money from clients. This is particularly dangerous for sex workers, who reported seeing their photos used to advertise services that they do not offer. This puts sex workers at risk for disagreements with clients who have seen or interacted with the catfish, increasing their chances of violence.
Most sex workers reported that their photos were either digitally scraped by bots or individuals with explicit intentions to gather catfishing materials. A common sentiment among sex workers who’ve experienced catfishing is that being digitally impersonated is almost inevitable when your work requires you to regularly post photos of yourself for marketing purposes. For those not willing to take on the constant vigilance required to protect their digital identity, giving up and accepting it is the only other option.
“A website which ‘advertises escorts’ copied and reposted a profile from another website which advertises my BDSM work (I am not an escort).” -Sex worker and AOP respondent
“Another sex worker stole my text and photos. Many advertising platforms have scraped my content and put it on their websites without asking but I have not had this happen on social media.” -Sex worker and AOP respondent
“I deliberately never used my face in photos to prevent identity theft anyway. I can’t stop catfishing from occurring, but I can mitigate its effect upon me.” -Sex worker and AOP respondent
“I gave up not sure there is anything to do, I don’t use my face online and even with
watermarks I know that’s a risk of doing this. I try to pretend like it doesn’t matter but its
made a difference for sure.” -Sex worker respondent
31.25% of those who reported being both digitally impersonated and simultaneously shadowbanned said that the catfishing account pretending to be them still showed up in searches. This phenomenon happened exclusively to those who identified as sex workers. This means that the platforms are invisibilizing the sex workers’ accounts while allowing the catfishing accounts to remain visible and searchable.
Some of these catfishing accounts do get reviewed by content moderators, but are not removed as they fall under the parameters of parody, art, or fan accounts. Parody accounts are typically made to mock, or “troll”, an intended user. Despite possibly constituting harassment, parody accounts are often difficult to have removed as parody is often allowed by platforms’ TOS.
One AOP reported that Twitter said in response to their report of being digitally impersonated that: “It wasn’t really catfishing, it was ‘parody’ and upheld as such by Twitter.” Another said: “[I was] trolled by a conservative hate monger account that posted my selfies without my consent. they coined me as a snowflake/’transtrender’/misandrist.”
43.75% of sex workers and 46.43% of sex workers who are also AOPs report having had their images or content used for a fake account that they did not run or provided false information (e.g. having a catfishing account steal their photos). AOPs who do not do sex work are significantly less likely to say the same (14.29%).
One sex worker who started working before they turned eighteen said: “I’ve had underage photos of me posted by other accounts. Now, these images are over 10 years old and I’m not sure who is posting them. I’ve also had agencies use my photos for bait and switch advertising.” Another sex worker said they constantly have digital impersonators: “[T]hose accounts never seem to be taken down despite multiple reports against them.”
“I have dozens of accounts using my photos and identity to scam money from people. Of course, we can’t get our account verified as sex workers so it’s difficult to get these accounts shut down.” -Sex worker respondent
Of those who were able to have their digital impersonator’s account removed, many had to have multiple people report the catfish profile to the platform in question. The process of reporting impersonation on social media platforms is often difficult to navigate, which adds an additional layer of inaccessibility.
Sex workers also reported having to provide social media platforms with government ID in order to prove they are the original account holder. This isn’t an option for those who do not have government ID and those who have a work name or pseudonym. Sex workers and AOPs are also hesitant to provide identification out of fear of legal action. While this danger has always been present for sex workers, the recent increase of protester arrests—made possible through social media identification—has increased the risks for everyone.
“Twitter asks you to prove that you are not the fake profile by providing them with evidence. But if you were in sex work and you use a fake name, how are you supposed to prove that?” -Sex worker respondent
“After getting a group of people to report said account, it was taken down. However I’m sure there are other people using my pictures to catfish.” -Sex worker respondent
“Didn’t bother flagging with Twitter as their TOS seem to imply that we’d need to show
ID to prove copyright ownership.” -Sex worker respondent
“Once I got myself deleted for reporting a catfish account. They were left to exist, and I was deleted.” -Sex worker respondent
In her project, We’re All Eating Catfish Tonight, sex worker, activist, and artist Mistress Wu processes the experience of having her images stolen to make catfish accounts. She writes a letter to each of these accounts: “I think that you desiring to be me is in a way the greatest compliment; it is saying that I am better than you; and yet in this digital realm, we are one and the same, made up of the same silicon, the same pixels as each other. I really do hope that I get to cross paths with you one day, in this body, or one of my 21 other bodies.”
Many sex workers face myriad barriers to marketing their services without access to the mainstream marketing channels and financial technologies marketers and small businesses are accustomed to. Marketing industry data confirms that the vast majority of small businesses have an overreliance on mainstream paid marketing channels—which sex workers can’t use—for early stage growth.
According to Statista, social media ad spending alone reached about $27 billion in the U.S. in 2018. A 2019 study found that search engine traffic generated 65% of total ecommerce sessions, 32% of which was generated through paid search ads. According to HubSpot, 68% of marketers say that paid advertising is “very important” or “extremely important” to their overall marketing strategy.
But that’s just the beginning.
Systematic Whorephobia in Marketing & Entrepreneurship
Awareness: While 92.45% of sex workers who work online use social media for sex work, they experience increased rates of:
- shadowbanning on major social media platforms (75% of sex workers have been shadowbanned; 85% on Twitter, 40% on Instagram, 10% on Facebook).
- having their posts removed from social media (66.67%).
- having content that does not violate a sensitive media policy marked as sensitive media (70.83%).
- being removed from social media platforms (deplatformed) altogether (32.65%). Sex workers who have been deplatformed from social media have been deplatformed more than 3 times on average, and 93.75% of sex workers who have been deplatformed do not get their accounts back.
Acquisition: Whether it’s web hosting or domain registration restrictions or post-purchase removal, sex workers can also be deplatformed from their “owned” marketing assets. Email marketing—another owned asset—generates $44 in revenue for every $1 spent.
Unfortunately, many email marketing tools—such as MailChimp—also restrict use for sex workers, leaving sex workers who have lists of engaged (potential) clients fewer tools to communicate with them (e.g. send newsletters, blog posts, promotional emails, etc.) The email tools that do not restrict sex workers have lower sender reputations because of their less rigid content moderation guidelines, meaning any emails sent are much more likely to be marked as spam in inboxes—and never read.
Activation: 49.06% of sex workers have had a payment blocked, cancelled, or flagged because of sex work. 66.67% have been completely deplatformed from financial technology (e.g. Stripe, PayPal, Venmo, etc.), which has impacted their ability to work and charge clients. Pushed off popular mainstream payment processors, sex workers are forced to use payment processors with much higher fees:
Retention: Sex workers can’t use Patreon, MailChimp, Facebook Ads, Google Ads—the list goes on. Recurring payment options are scarce, so each repeat purchase is a conscious, labor-intensive decision. The majority of methods clients have for contacting a sex worker they’ve previously engaged with are subject to content moderation and deplatforming as well—save for encrypted services, which are also under attack by legislation like the EARN IT Act.
Get That Bread, Get That Head, Then Leave. Peace Out.
Many platforms have grown due to early adoption and innovation from sex workers. After achieving mainstream success, this is often swiftly followed by sex workers’ removal from the platform. Tumblr is a classic example of this pattern.
In 2013, Tumblr was acquired for $1.1 billion USD. 5 years later, in December 2018, Tumblr made the decision to ban adult content. The platform has been on a steady decline in visits since then. In August 2019, Tumblr was purchased for less than $3 million USD, 0.3% of the 2013 valuation.
Now we see OnlyFans on a similar trajectory. The platform was launched in 2016 and rose to mainstream popularity by marketing as a stable marketing platform for sex workers. As of May 2020, the site has 24 million registered users and claims to have paid out $725 million to its 450,000 content creators. Tim Stokely, CEO of OnlyFans, said, “The site is seeing about 200,000 new users every 24 hours and 7,000 to 8,000 new creators joining every day”, demonstrating exponential recent growth (1.78% creator growth every day).
Beyoncé mentioned OnlyFans in her verse on Megan Thee Stallion’s “Savage Remix”, which resulted in an immediate 15% spike in traffic. Blac Chyna, Cardi B, Bella Thorne and other celebrities have created OnlyFans creator accounts themselves. Thorne’s OnlyFans debut earned her around $1.3 million in her first 24 hours—just another example of how mainstream non-sex workers appropriate sex worker aesthetics and platforms for their own financial gain.
As the platform picks up mainstream appeal, sex workers are already starting to feel the effects of increased content moderation, including not being able to mention in-person sex work on OnlyFans. Some sex workers have even reported deplatforming, a sign history is repeating itself and they will be penalized and/or completely erased after the platform reaches a certain point of growth.
Between evolving and outright disappearing platforms, like Tumblr and OnlyFans, some sex workers are still managing to build thriving small businesses online without access to the world’s most popular marketing channels and financial technologies. According to BuiltWith, 50.21% of the top 10K sites use Facebook Custom Audiences. 40.03% of the top 10K sites use Google Remarketing. 14.38% of the top 10K sites use Twitter Ads. Sex workers are unable to use any of these channels. In fact, 17.08% of the top 10K sites use Stripe, which also discriminates against sex workers, making it one of the most popular payment processors available today.
Tumblr and OnlyFans would not have the growth trajectories they do without access to paid social ads, paid search ads, organic social media, email marketing tools, etc. Nor would the marketing channels themselves, like Facebook and Twitter. Similarly, we must ask ourselves what sex working small businesses’ growth trajectories might have been with these marketing channels.
Whether you’re looking at social media platforms, website hosting, financial technologies, the world of small business limits—if not excludes—sex workers at every stage of the traditional marketing funnel. Sex workers continue to do their best to build thriving small businesses despite this barrage of marketing barriers. With many sex workers funding their movement work through sex work, this has a compounding negative effect on the sustainability of both community organizing and sex worker’s ability to make a living (to support themselves and their families).
Discussions and Recommendations
Our research suggests that many sex workers and AOPs use platforms and technologies to work safer, build community, and earn a living. When these platforms become unusable to communities vulnerable to state and platform policing, it: exacerbates inequalities, disrupts movements, chills speech, isolates communities, and decreases ability to make and redistribute money.
Barring sex workers from financial technologies and social media is a form of structural violence. Content moderation practices reflect the biases of the creators of the platforms and algorithms, as well as the biases of the content moderators themselves. Any laws or content moderation practices that hinder an individual’s access to technology and resources will only exacerbate harm.
Here are four harm reduction recommendations for platforms:
- Make internal content moderation practices public. There should not be a difference in the rules, TOS, or community guidelines provided to users and content moderators.
- Give users more choice in what they see. Instead of relying on the labor of content moderators, overly broad content moderation algorithms, and automatic settings that restrict what users see, users should be able to determine for themselves what content should be prioritized and/or hidden from view.
- Hire sex workers to conduct compentency trainings for staff. The inclusion of sex workers in the development of technology needs to be thought of as a diversity issue. Academics, journalists, legislators, and people who create technology should not build it without consulting the communities that are most impacted by their creations. The only way to ensure that new and emerging technologies don’t cause more harm is by listening to the expertise of the communities most impacted by poorly designed technology, legislation, and infrastructure.
- Tech is not neutral. Tech policy is not neutral. Silence is not neutral. All tech is political, all tech impacts human rights in a variety of interconnected ways. If you work in tech, open your wallet and share this report with colleagues. It is your personal responsibility to make everyday decisions that reduce the harm tech causes to marginalized communities.
- Listen to sex workers when they warn about the potential impact of policies. Sex workers expertly predicted the impact of FOSTA. Sex workers’ vast knowledge and expertise cannot be ignored. This report shows how losing access to technologies or platforms—or even just reduced visibility—pushes people into increased financial precarity.
- Stop the state from determining what safety means for communities by stoking fear and spreading misinformation. Legislation is being signed into law with language to ostensibly stop human trafficking. For example, the EARN IT Act is being pushed through as a way to end CSAM. We cannot allow the state to sign legislation into law that impacts everyone’s privacy, security, and safety.
- Challenge the framing that sex workers and survivors are two discrete communities. Sex workers trade sex for myriad reasons and under myriad different economic situations. Whether through choice, circumstance, or coercion, sex workers need labor rights and access to the same working tools as their civilian peers. We do NOT need legislation that provides a few survivors legal recourse after the fact.
- Fight against legislation that increases liability for platforms and does nothing to stop violence. Legislation threatening § 230 is being routinely pushed through under the guise of supporting survivors. This rhetoric has been effective for gaining political support, but does nothing to actually stop violence before it happens. We’re left with legislation that increases civil and criminal liability for platforms while only providing legal recourse for a very select few survivors.
- Advocate for the decriminalization, decarceration, and destigmatization of sex work. As per DecrimNY, we call for policy makers to:
- “Decriminalize sex trade related offenses in New York that harm people who do sexual labor by choice, circumstance, or coercion, including sex workers and people profiled as sex workers, as well as people who purchase sexual services. Pass legislation and implement administrative policies that protect people in the sex trades from economic exploitation as well as interpersonal violence.
- Decarcerate people who have been arrested on sex trade-related offenses so that people can move forward with their lives without lingering ties to the criminal legal system. Vacate criminal records related to prostitution and end the ongoing entanglement with the court system that the rescue industry produces.
- Destigmatize the sex trade so that workers have access to housing, education, employment, health care, and other basic needs without restriction. Not everyone trading sex wants to continue doing so and we support evidence-based, harm reduction-rooted policies, and funding that supports people’s safety and empowers those seeking different work.”
The primary survey consisted of a mix of quantitative and qualitative questions. The survey was disseminated online through social media platforms. In order to complete the survey, individuals had to have some sort of access to the Internet. We did not collect data on respondents’ race or gender, leaving a need for further research on how the content moderation of sex workers operates within racist, sexist logics. We suggest reading Salty’s report on Algorithmic Bias to learn more about this.
When analyzing the data on shadowbanning, we only analyzed the results of people who were familiar with the term. This means that people who were unfamiliar with the term were left out of the sample, whether or not they had experienced shadowbanning. As shadowbanning is an opaque practice, it is likely that more individuals have been shadowbanned than those who were familiar with the term.
A certain political awareness is required in order to identify your experiences of reduced visibility on social media as “shadowbanning”. A lot of people, including some sex workers and AOPs, think that the failures of tech are user error or glitches rather than realizing they have been deliberately targeted. However, people who self-identify online as “sex workers” (a political term in itself) and “AOPs” are both more likely to understand these experiences as shared among people who similarly self-identify, making them more likely to feel specifically targeted by platforms.
Our research looks exclusively at self-reported user experiences. We also did not have a non-sex working and non-AOP comparision sample. Further research is needed to understand how content moderation practices are felt by the general public. Our research shows that sex workers feel the impact of content moderation practices more severely than AOPs who do not trade sex. It is likely that when compared with the general public, this difference would be even more dramatic.
With COVID-19 and uprisings against police violence occurring simultaneously globally, we are studying an ecosystem that may not be representative of the rest of the year. Or, for that matter, the rest of the decade. What we do see in the data is how individuals are harmed by negligent content moderation practices.
As we write this report, content moderation practices are rapidly changing. As is the public’s awareness of shadowbanning practices. Algorithms used by platforms to moderate content online change and are updated at a rapid pace. Since starting this report, we have had to update it multiple times to reflect the growing knowledge and changing platform admissions of shadowbanning.
This research focuses on sex workers and AOPs who use social media for their work. Much of sex work and activism happens offline. More research is needed to better understand how the discrimination and policing of these platforms follows people offline, and how social media companies collaborate with state actors to aid in the deportation and incarceration of marginalized communities. Further research would benefit from asking what type of organizing individuals do and how they financially support their organizing.
This research explores the relationship between sex workers, AOPs, and the digital platforms they use to work and organize. More research is needed to explore the common relationship of the intersecting identity of sex worker and AOP. Many AOPs fund their unpaid labor through their work in the sex trades, and more research is needed to fully understand this relationship and how platform punishment impacts those who support their activism work with sex work.
In Hacking//Hustling’s past research, we found that 50% of online sex workers reported having physical disabilities or mental health issues that prevented them from engaging in traditional forms of labor. Because of the harsh form of content moderation faced by sex workers on social media, we are forced to adapt and alter the ways that we speak online to avoid platform punishment.
As of the publication of this report, algorithms have not yet caught up to fully moderating text in images. Sex workers often use text in images or intentional misspellings of words while sharing information with clients and community to avoid platform punishment.
Sex workers are unable to use accurate alt text on our images for fear of shadowbanning and deplatformming. This makes content and resources inaccessible for screen readers used by blind people. Sex workers are forced into censoring themselves in a way that makes their content and harm reduction materials illegible to our disabled comrades, further facilitating harm.
Future research should continue to explore the relationship between sex work, disability, and online work. Including how sex workers’ forced self-censorship and use of steganography makes harm reduction materials less accessible to our blind, hard of seeing, or otherwise disabled community members.
Despite platforms’ ubiquitous denial of shadowbanning as a practice, this report confirms both the existence and harmful impacts of these opaque content moderation practices. This research on content moderation follows, and acts as an extension of, Hacking//Hustling’s previous report, Erased.
With the EARN IT Act, a variety of anti-encryption legislative efforts, and the defunding of open-source technologies on the horizon, this report further highlights the harms to—and complexities of—sex workers and AOPs who are trying to use social media to make a living, share resources, and organize.
Our research begins to show how the identities of sex workers and AOPs intersect to create more severe content moderation for people who hold both identities. The more you use a platform for your activism or your sex work, the more likely you are to have your content repressed or invisibilized by algorithms.
The paradox of the seemingly unyielding surveillance and censorship of digital technologies coupled with our growing reliance on these technologies has brought about another example of sex workers and AOPs serving as canaries in the coal mine—suffering the extent of these restrictions before they are noticed by the general public. The existence of shadowbanning has been slowly pieced together by social media users who notice changes to the platforms and in the algorithms. These changes indicate a consistent and deliberate use of content moderation tactics meant to suppress content from, and demobilize, certain communities.
Adult content guidelines have been used as an excuse for the systematic violence that denies sex workers free access to social media and financial technologies, which are both a means to more safely make a living. Now, the same tactics that are used to repress sex worker voices are more visibly being used to suppress the Black Lives Matter movement. The recent uprisings against racism and police brutality have brought civil rights activists and organizers into an increased awareness of how their content is being monitored, policed, and suppressed.
While platforms still deny the practice of shadowbanning, communities who experience harsh forms of platform policing have shared in this report how harmful content moderation practices interfere with their ability to make a living and organize powerful movement work. We call for platforms and policy makers to act on the recommendations resulting from this extensive research.
Important Terms and Concepts
What are AOP?
AOP is an abbreviation of activists, organizers, protestors that we use in this report for brevity and clarity.
What is Section 230?
Section 230 of the Communications and Decency Act (CDA) is the most important and referenced provision, which states that: “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” This provision means that Internet platforms are not liable for the content users produce and post to their platforms. For example, Yelp is not liable for defamation if a Yelp user posts a negative restaurant review, even if the post is malicious. Section 230 also contains a “safe harbor” clause, which states that providers and platforms don’t lose their 230 protections if they moderate their websites (e.g. remove offensive or harmful content).
Section 230 was written in response to seemingly perverse legal outcomes in a series of cases decided in the 1990s which resulted in websites having more legal liability if they took it upon themselves to moderate offensive user-generated content than if they didn’t moderate content at all. Many have said Section 230 is responsible for the creation of a free and open Internet. Without it, the Internet might have looked much more like traditional publishing or other forms of media, in which content is made only by a small group of people.
(This definition was adapted from SurviveEARNIT.com.)
What is FOSTA?
Public Law 115-164, better known as FOSTA-SESTA (“Fight Online Sex Trafficking Act” and “Stop Enabling Sex Traffickers Act”), became law on April 11th, 2018, with bi-partisan support. FOSTA was the first substantive amendment to Section 230, creating a broad carveout for civil lawsuits against Internet providers and platforms.
FOSTA also expanded criminal liability for any Internet platform, including apps and listservs—which facilitate prostitution, essentially creating a federal third-party law with a maximum penalty of up to 25 years in prison. The stated goal of this law was to reduce human trafficking by amending Section 230 of the Communications Decency Act.
The law has actually put increased pressure on Internet platforms to censor their users and push vulnerable communities into increased financial insecurity, housing instability, and exposure to violence. In July 2020, CityXGuide was seized and its owner was arrested, serving as the first use of the new criminal provisions created by FOSTA.
(This definition was adapted from SurviveEARNIT.com.)
What is the EARN IT Act?
The EARN IT Act is a bill that, as of the publication of this report, has passed the Senate Judiciary Committee. The Act proposes a 19-member commission tasked with creating best practices for tech companies in addressing all elements of child exploitation.. It is unclear how these best practices will be used, as in the bill’s current draft they are non-mandatory recommendations.
The EARN IT Act creates a carveout to 47 U.S.C. § 230 as it pertains to child exploitation laws. This carveout is similar to the carveout of § 230 created by FOSTA-SESTA in 2018, but is more broad. Section 230 has never provided immunity to tech platforms that host content that violates federal criminal law. EARN IT specifies that tech platforms will additionally have no § 230 immunity for hosting content that violates : (1) federal civil child exploitation law 18 U.S.C. § 2255 (civil claims related to 14 federal criminal laws relating to child exploitation); (2) any state criminal law relating to child sexual abuse material (CSAM) that uses the definition of CSAM found in 18 U.S.C. 2256(8); or (3) any state level civil claim related to CSAM that uses the definition of CSAM found in 18 U.S.C. 2256(8).
The likely impact of EARN IT’s § 230 carveout, if passed into law, is that tech companies will adhere to the standards created by the most conservative states and that sexuality- and sexual health-related content—particularly content aimed at educating youth—will be even further suppressed online. To read more about the EARN IT Act see this section-by-section breakdown.
(This definition was adapted from SurviveEARNIT.com.)
What is the PATRIOT Act?
was passed in 2001, just after 9/11 as part of U.S. anti-terrorism initiatives. This law increased the surveillance powers of the U.S. Government with the stated goal to make it easier for state and federal agencies to share information.
What is “white slavery”?
A term used by British and American AOPs, journalists, and politicians to describe an imagined epidemic of forced sex work at the turn of the 20th century, in which fears of industrialization, new technologies, and miscegenation manifested as public narratives of white women and girls being lured or kidnapped into prostitution, usually by African American or immigrant men. “White slavery” rhetoric resulted in the passage of the “White Slave Traffic Act of 1910,” also called the Mann Act, which is still in use today and which was expanded in 2018 by FOSTA.
What is content moderation?
In this research, we are using the definition of content moderation as introduced by Sarah T. Roberts: “content moderation is the organized practice of screening user-generated content (UGC) posted to Internet sites, social media and other online outlets, in order to determine the appropriateness of the content for a given site, locality, or jurisdiction. The process can result in UGC being removed by a moderator, acting as an agent of the platform or site in question. Increasingly, social media platforms rely on massive quantities of UGC data to populate them and to drive user engagement; with that increase has come the need for platforms and sites to enforce their rules and relevant or applicable laws, as the posting of inappropriate content is considered a major source of liability.”
What is an algorithm on social media?
The algorithm in this context refers to the machine learning algorithm deployed to stack-rank the content a user is shown in Feeds and Stories. The algorithm predicts what content is most important to you and shows you this content. Notably, it also does this with paid advertisements. The algorithm considers what type of content it is, who it’s from, hashtags used, if it’s an original post or a repost, etc. It is impossible to know every single factor that goes into the algorithm, as it is constantly adapting and changing itself and the algorithm can also be changed and adjusted manually by the company over time.
What are blackbox technologies?
Blackbox technologies can be understood through the inputs and outputs created, without human knowledge of the internal workings. For example, many algorithms are blackboxes and even the platforms who use them may not be entirely sure of the full extent of their workings.
What are TOS?
TOS is short for Terms of Service. TOS are a contract between a platform and a user. TOS are notoriously and intentionally difficult to read and inaccessible. Some research shows that anywhere between 91-97% of users do not read TOS before signing them.
What are community guidelines/standards?
Community guidelines are documents developed to identify content or behavior that is not allowed on a given platform. Often, these documents are incorporated by reference into terms of service, but are meant to be much more readable and less realistic. However, they usually are very high level, leaving unclear exactly what content is allowed or not allowed on the platform. Many platforms have internal documents that lay out in more detail how to enforce their rules.
What is platform policing?
Platform policing is defined, in this research, as the set of content moderation tools used by platforms to enforce a sanitizing “universal” content standard as defined by platform designers, which invisibilize marginalized communities, quells free speech, and demobilizes movement work. This can include shadowbanning, deplatforming, cross-platform punishment, and platforms actively collaborating with the carceral state.
What is shadowbanning?
Shadowbanning is a tool that platforms use to reduce the prevalence of content that the platform deems “high-risk” and that should not be easily discoverable. Typically, when a user is looking for content or profiles on a platform it is easily found through searches or suggestions. Shadowbanning takes a few forms that makes visibility and discovery more difficult: content not showing up on the discover tab when you are searching for a profile, content not showing up in relevant hashtag pages, and profiles not showing up when searched for in the search bar, unless using a full screen name spelled correctly and clicking enter. Because of these types of reduced visibility and discoverability, an account might show up less in other users’ feeds, unable to connect with new followers. At times, shadowbanning can make social media platforms unusable, for example, when you are unable to connect with or find community and clients.
What is invisibilization?
Invisibilization is the process by which platforms render your social media account invisible to other users. Invisibilization can be a product of shadowbanning or other opaque content moderation practices.
What is deplatforming?
Deplatforming is best known as a type of political activism that intends to shut down the platforms of controversial speakers and deny them access to sharing their beliefs in public spaces. Most notably, deplatforming is used to talk about the platform response to removing the platform or monetization methods of prominent alt-right speakers.
We are intentional about our use of the word deplatforming to discuss the ways that sex workers are systematically removed from platforms because sex workers’ removal and invisibilization on social media and financial platforms are a reflection of the ideologies of the platforms that deem sex workers as “high-risk” and “controversial”.
It is also important to note that sex workers are often deplatformed at a significantly faster rate than prominent figures or media sources of the alt-right. For example, Cloudflare was quick to remove Switter, but has been notoriously slow to remove alt-right websites. Any tools or social policies that are used to limit the spread of alt-right ideologies are also used to police marginalized communities, often in a harsher and faster way.
The accelerated rate that sex workers are deplatformed and the labeling of sex workers as higher risk can be understood through the lens of criminalization and legal liability: sex work is criminalized, but nazism is not.
What is structural gaslighting?
Structural Gaslighting is when the state, structures or or institutions deny a set of practices which certain users or communities know to be true. In this report we talk about policing and platform policing tactics that are denied by those in power to be forms of structural gaslighting. In the Scientific American, authors of George Floyd’s Autopsy and the Structural Gaslighting of America provide a powerful example of structural gaslighting and say, “Black people are suffocating under the weight of anti-Black hatred. They cannot breathe. And even as they gasp for air, structural gaslighting operates to deny the truths of the causes of their suffocation.”
What is platform gaslighting?
Platform gaslighting is a concept that we use in this report to describe the structural gaslighting that occurs when platforms deny a set of practices which certain users know to be true. Gaslighting is defined as psychologically manipulating someone into questioning their own sanity. When platforms deny something like shadowbanning and users feel the impact of it, it creates an environment in which the shadowbanned user is made to feel crazy, as their reality is being denied publicly and repetitively by the platform.
What is networked shadowbanning or networked deplatforming?
In this research we propose the idea of networked shadowbanning and networked deplatforming to discuss the occurrence of a user’s “high-risk” digital footprint following them cross-platform or into the physical world. For example, the opaque practice reported by sex working community of bots scraping escort ads for e-mail addresses or metadata and these lists being shared between platforms to facilitate cross-platform punishment and/or deplatforming. The punitive measures of the platform can even follow individuals into physical spaces when social media platforms share information with the police or people lose access to income or financial technologies. It can go cross-platform or through social circles/communities.
What is the compounding effect?
In this report, this term is used to describe the magnified impact of platform policing on individuals who identify as BOTH a sex worker and an AOP as the compounding effect of platform policing.
What is catfishing?
Catfishing is the act of creating an online persona by posting photos or videos of another person, with the intention of interacting with others while pretending to be the person pictured.
What is fintech?
Fintech is short for financial technology. It encompasses technologies and computer programs that either facilitate or disrupt banking and the movement of capital online. Fintech includes crypto-currencies, crowdfunding, budgeting apps, and subscription services. The findings of this report focus on payment platforms, such as Venmo, Stripe, PayPal, Square Cash.
What is a hashtag?
Wikipedia defines a hashtag as “introduced by the number sign, or hash symbol, #, is a type of metadata tag used on social networks such as Twitter and other microblogging services. It lets users apply dynamic, user-generated tagging that helps other users easily find messages with a specific theme or content.”
What is a bot?
Dictionary.com defines a bot as “a software program that can execute commands, reply to messages, or perform routine tasks, as online searches, either automatically or with minimal human intervention.” Social media platforms demote and suspend bots and accounts that look like bot-like behavior. Many sex workers report less platform punishment after turning off automatic tweets, which is flagged as bot-like behavior by the platform, from 3rd party sites notifying customers of sales.
What is scraping?
Scraping or “web scraping” is the often automated process of collecting and extracting large amounts of data from a website. This may involve downloading individual web pages, entire websites, or email addresses of sex workers advertising on a platform.
What is surveillance capitalism?
“Surveillance capitalism describes a market driven process where the commodity for sale is your personal data, and the capture and production of this data relies on mass surveillance of the internet.” This term was coined by Shoshanna Zuboff in 2014.