Unbowed - Tracking technology-facilitated gender-based violence

Law & Policy Platforms & AI Courts Research Regions About us

Author: Staff reporter

  • Forty-eight US states now have laws on sexually explicit deepfakes

    Forty-eight US states now have laws on sexually explicit deepfakes

    Forty-eight American states have now enacted laws covering the creation or distribution of sexually explicit deepfakes, according to the third annual survey of state legislation by Ballotpedia, a non-profit that tracks US politics and policy.

    The count rose from 46 in January to 48 by 23 July, the cut-off for the report. Only Ohio and New Mexico have no such law.

    What those 48 laws do is not uniform. Ballotpedia describes them as laws “concerning the creation or distribution of deepfakes that depict explicit sexual acts or other sensitive content”, and notes that some address only the creation and distribution of child sexual abuse material, while others cover the non-consensual creation and distribution of adult intimate images. A woman whose image is used to make a sexual deepfake is therefore not equally protected across the 48.

    The two gaps

    Ohio has enacted no deepfake legislation of any kind — neither on sexual material nor on political communications. Its Senate passed Senate Bill 163 unanimously on 20 May, criminalising the making, sending or possession of AI-generated child sexual abuse material and requiring AI systems to watermark the images they produce. It went to the House and has not become law. It does not cover adults.

    New Mexico’s position is different. It is one of 33 states with a law regulating deepfakes in political advertising, but it has not legislated on sexual ones. It has addressed the use of synthetic media against candidates, and not against women.

    Neither state is without any law. All 50 states and the District of Columbia prohibit the distribution or production of non-consensual intimate images, according to Ballotpedia’s separate tracking, current as of June. The gap in Ohio and New Mexico is specifically about material that has been generated or altered by artificial intelligence.

    Federal law also reaches it. The Take It Down Act, signed in May 2025, criminalises the non-consensual publication of intimate images including deepfakes in some circumstances, and required covered platforms to put removal processes in place by 19 May this year.

    The first person convicted under that act was from Ohio. James Strahler II, 37, pleaded guilty on 7 April to cyberstalking, producing child sexual abuse material and publishing digital forgeries — the statute’s term for deepfakes. The Justice Department said he had used images of boys he knew to create sexual material of them. In May it announced the arrests of two men over albums said to depict about 140 female victims, and in June it seized two websites that distributed AI-generated nude images of women.

    Slowing down

    State legislating on deepfakes is losing pace. Lawmakers enacted 58 deepfake-related bills of all kinds in 2026 up to 23 July, against 64 by the same date in 2025. Almost half of this year’s enacted bills had sponsors from both parties.

    Laws on political deepfakes grew faster this year than sexual ones, from 28 states in January to 33 in July. Those laws have run into the First Amendment in a way the sexual ones have not: Senior US District Judge John Mendez blocked enforcement of California’s AB 2839 in October 2024, and Ballotpedia records three states whose pre-election prohibitions carry no exemption for material that discloses it is synthetic.

    Ballotpedia tracks bills, not prosecutions, and no state publishes figures on charges brought under its deepfake laws. Minnesota’s ban on nudification tools, which took effect on 1 August after a federal judge refused Elon Musk’s xAI an order pausing it, is among the newest and the most directly aimed at the tools themselves rather than the people who use them.

  • Nudify app ads still appearing on Instagram, campaigners say

    Nudify app ads still appearing on Instagram, campaigners say

    Advertisements for artificial intelligence tools that generate sexual images of real people without their consent are still running on Instagram, the campaign group Stop Online Harm said on Wednesday.

    The organisation said its team had investigated and found ads promoting tools that digitally remove clothing and produce sexually explicit images. The adverts it found offered to “undress” the subjects of photographs and showed women in sexually explicit positions, it said.

    Meta, which owns Instagram and Facebook, prohibits exactly this. Its Adult Sexual Exploitation community standard bars “promoting, threatening to share, or offering to make NCII or kissing imagery through applications or services (including AI), or instructions, even if there is no commercial or non-commercial imagery shared in the content (sometimes referred to as ‘nudify’ apps or ‘AI kissing’ apps)”. NCII stands for non-consensual intimate imagery.

    Stop Online Harm published its finding on LinkedIn.

    What has been documented

    A detailed account of nudify advertising on Meta’s platforms was published on 26 July by the Tech Transparency Project, a Washington-based non-profit that monitors large technology companies.

    Using Meta’s Ad Library — the disclosure database the company is required to maintain in the European Union under the Digital Services Act — the group traced a network of adverts to GatherOne, also trading as Hongkong Gather Wisdom, one of roughly a dozen Chinese agencies Meta authorises to place adverts on behalf of Chinese clients. Because China blocks both platforms, Chinese advertisers wanting to reach audiences abroad must in principle go through such an intermediary.

    The group identified 210 Facebook pages naming GatherOne in their advert disclosures, which between them ran about 30,800 adverts for face-swap and other AI apps. Of those pages, 43 ran more than 7,600 adverts for apps the researchers verified could nudify women. Multiple user reviews of one app flagged what appeared to be child sexual abuse material.

    Meta removed many of the adverts and disabled accounts during the investigation, the group said, but at least two dozen linked pages carried on advertising after having adverts taken down.

    Meta did not answer the Tech Transparency Project’s questions. In a statement to Bloomberg News, which reported the research, it said it takes “aggressive steps” against nudify apps and that advertisers who break its rules risk having adverts cancelled, being fined and having their accounts terminated. GatherOne told the researchers after publication that it had suspended new advertising accounts for AI nudify, face-swap and “girlfriend” services.

    Meta has not commented publicly on Stop Online Harm’s finding.

    Not the first time

    Meta sued Joy Timeline HK, the Hong Kong company behind the CrushAI nudify app, in June 2025, saying it had repeatedly evaded enforcement. Researchers had linked more than 87,000 adverts across over 135 Facebook pages to that app. A CBS News investigation separately found hundreds of nudify adverts running on Facebook and Instagram.

    Creating these images is now unlawful in a growing number of jurisdictions. Minnesota’s ban on tools that generate fake nude images took effect on 1 August after a federal judge refused xAI’s request to pause it.

    A poll published on 4 August by Women Who Tech and Rad Campaign, two organisations that campaign on the issue, found 67 per cent of American adults registered some or high concern about AI being used to create sexual imagery of women, and 71 per cent about deepfakes that sexualise children.

  • Nearly nine in ten Australian women in politics report online abuse

    Nearly nine in ten Australian women in politics report online abuse

    Nearly 90 per cent of Australian women who hold elected office or are seeking it have experienced online abuse or intimidation linked to their political participation, according to a survey published on Wednesday by Women for Election, an organisation that trains women to stand for public office.

    The survey is the organisation’s own research, commissioned in July 2026.

    It is one of the few Australian datasets that asks women in politics directly about AI-generated abuse. Sixty-five per cent of respondents said they had personally experienced or witnessed AI-generated abuse, deepfakes or manipulated content targeting women in public life, and 98 per cent said they were concerned artificial intelligence would make the problem worse over the next five years.

    Half now question whether they should have stood

    Eighty-two per cent of respondents said they had changed their behaviour because of online abuse. Of those, 58 per cent limited their use of social media, 36 per cent increased security measures and 34 per cent took time away from campaigning or public life. Half said that, knowing what they now know, they were unsure they would run again or would not have run at all. Sixty-four per cent said abuse had made them less likely to stand again, seek higher office or encourage other women to stand.

    Forty-five per cent said family members or campaign supporters had been targeted because of their political involvement.

    Only 47 per cent of those who experienced abuse reported it, and of those, 69 per cent were dissatisfied with the response they received. Women for Election’s media release states that 110 women were surveyed.

    Helen Blackburn, mayor of Bundaberg Regional Council in Queensland, told Australian Associated Press she had faced a campaign of AI-generated fakes across several platforms in the past two months, including a parody series made after she announced her council’s budget. One item featured the Grim Reaper, she said, followed by users writing that they hoped she would die and that “all I need is two bullets and I’ll sort it out”. Blackburn said she had considered resigning.

    The global picture

    Women for Election sets its findings alongside research led by Julie Posetti of City St George’s, University of London, for UN Women. That study, Tipping point: The chilling escalation of online violence against women in the public sphere, published in December 2025 and drawn from a 2025 survey of women journalists, human rights defenders, activists and other public communicators, found that 70 per cent had experienced online violence in the course of their work and that more than 40 per cent reported offline harm linked to online abuse.

    Among women journalists and media workers specifically, the share who linked offline attacks, abuse or harassment to online violence rose from 20 per cent in a 2020 UNESCO survey, also led by Posetti, to 42 per cent in 2025 — more than double.

    Women for Election is calling for greater accountability from social media and AI companies, dedicated protections for candidates during election periods, and a national action plan.

    Australia already has a statutory route. The eSafety Commissioner can require platforms to remove adult cyber abuse, but the threshold is narrow: the material must target one particular Australian adult, must be menacing, harassing or offensive, and must meet a test of serious harm judged by what an ordinary reasonable person would conclude rather than by the effect on the person targeted.

  • US poll shows huge concern over AI sexual images of women

    US poll shows huge concern over AI sexual images of women

    Two-thirds of American adults say they are concerned about artificial intelligence being used to create sexual imagery of women, according to a poll its authors describe as the first to put a number on public opinion about it.

    The survey found 67 per cent registered some or high concern about AI being used to create sexual imagery of women, and 71 per cent about AI deepfakes that sexualise children. It was published on 4 August by Women Who Tech, a non-profit network of women-led startups, and Rad Campaign, a digital agency that works for non-profits.

    Both organisations campaign on this issue and the poll is their own.

    Allyson Kapin, founder of Women Who Tech and a founding partner at Rad Campaign, said the results showed “policymakers and Big Tech companies are lagging far behind their own constituents and customers on this issue”. She said there was “a clear bipartisan majority for safeguards against sexual deepfakes of women and children”, adding: “Victims have waited long enough.”

    Respondents who reported “some” concern are counted alongside those reporting “high” concern, so the headline numbers measure the breadth of unease.

    Fieldwork ran from 13 to 18 June 2026, seven weeks before publication, among 1,015 adults aged 18 and over. Responses were collected online and weighted. The authors give a Bayesian confidence interval of 3.5 for 1,000 interviews, which they describe as roughly equivalent to a margin of error of plus or minus 3.1 points at 95 per cent confidence.

    The survey followed the period in which X was flooded with AI-generated sexual images produced by Grok, the chatbot built by Elon Musk’s xAI, depicting mostly women and children.

    Two bills are waiting in the House of Representatives. The DEFIANCE Act, which would allow victims of sexual deepfakes to sue those who make and share them, passed the Senate unanimously in January. The ENFORCE Act, which would let prosecutors charge AI-generated child sexual abuse material as they would any other, has also cleared the Senate.

    The TAKE IT DOWN Act, which requires platforms to remove non-consensual intimate images within 48 hours of a valid request, took effect on 19 May. The poll’s authors say it is too early to judge its effect.

  • Tokyo police arrest man over AI nude images of women athletes

    Tokyo police arrest man over AI nude images of women athletes

    Tokyo Metropolitan Police arrested a 32-year-old company employee on 3 August 2026 on suspicion of using generative artificial intelligence to turn photographs of real women into sexually explicit images and posting them on social media.

    The man, from Himeji in Hyōgo Prefecture, was arrested on suspicion of defamation and of violating Japan’s Act on Punishment of Activities Relating to Child Prostitution and Child Pornography. He has not been charged. Under Japanese procedure, prosecutors decide separately whether to indict.

    Police allege that in February he took photographs of three women in their twenties wearing the athletics uniforms they had worn at high school, used generative AI to alter them into nude images, and posted them where anyone could see them. He is reported to have paid a subscription of around 3,000 yen a month to a generative AI website to produce them.

    Yomiuri Shimbun reported that he created about 300 such images and posted some of them. The broadcaster TBS reported that around 3,800 obscene images and videos were found on his phone, and that police are investigating further offences. The two figures come from different outlets and we have not reconciled them.

    The second allegation is the legally unusual one. Police say he also altered a photograph of a woman now in her forties, taken at a school sports day when she was a junior high school pupil, at the request of a 17-year-old boy in Kagoshima Prefecture. Because the original photograph showed a child, child pornography law applies, even though the woman depicted is now an adult. The teenager has been referred to prosecutors.

    That distinction matters beyond this case. It means the age of the person in the source image, not the age of the person today, determines which law reaches the offence.

    Japan has no dedicated offence covering non-consensual intimate images of adults. That gap explains the reliance on defamation for the first set of allegations, a charge that turns on damage to reputation rather than on the violation itself.

    The case also points to a pattern documented elsewhere: sportswomen targeted because of what they wear. In a remark reported by several Japanese outlets, the man said athletics uniforms were easy to process.

    Japan’s National Police Agency recorded 123 reports and consultations about sexual deepfake images of minors between January and June 2026, exceeding the 114 logged across the whole of 2025.

  • Bangladesh sets out zero-tolerance policy on online abuse of women

    Bangladesh sets out zero-tolerance policy on online abuse of women

    The Bangladeshi government has restated a policy of zero tolerance towards gender-based and online violence, identifying digital safety as a priority as internet use expands.

    Measures described include enacting and enforcing laws against cyber violence and sexual harassment, faster investigation and prosecution of offenders, and specialised institutional machinery including a dedicated intelligence unit and a cybercrime tracking system.

    This has been set against cases that campaigners say demonstrate the distance between policy and practice.

    An opinion article published in the Daily Star on 29 July 2026 described a case in Bagerhat in which a woman was recorded without her knowledge and then blackmailed. When she refused to comply, the article said, the man sent altered footage to her husband. She was beaten and later divorced. The article said her tormentor was never brought to justice.

    Unbowed has not been able to corroborate that account.

    Figures circulating in Bangladeshi media put the proportion of women social media users who have experienced online violence at 89 per cent, and the number of women seeking help over cybercrime in 2024 at more than 60,000. Neither figure has been traced to a published methodology.

    Technology-facilitated gender-based violence — abuse committed, assisted or amplified by digital tools and directed at people because of their gender — has also featured in Bangladeshi electoral politics, with women candidates reporting organised online harassment. Women make up about 4 per cent of candidates contesting, and 30 of 51 parties registered for the elections fielded no women at all.

  • Indian police treat coordinated online abuse as sexual harassment

    Indian police treat coordinated online abuse as sexual harassment

    Police in the Indian city of Bengaluru have opened an investigation into coordinated online abuse under provisions covering sexual harassment, after a complaint from the actor and former member of parliament Ramya.

    A first information report was registered on 28 July 2026 naming 43 social media accounts. The complaint alleges the accounts are operated by supporters of another actor, Darshan Thoogudeepa, who is not accused of any wrongdoing. The case has been transferred to the Central Crime Branch of the city police for a detailed investigation.

    Online abuse in India is more commonly pursued under general provisions of the Information Technology Act, which address individual acts by individual users. Invoking sexual harassment provisions against a group of accounts treats the campaign itself as the offence.

    Whether that holds is the question the case will settle. Coordinated abuse is difficult to prosecute because responsibility is distributed: each account may post material that would not, alone, meet a threshold, while the aggregate effect on the person targeted is severe. Few jurisdictions have found a way to reach it.

    Technology-facilitated gender-based violence — abuse committed, assisted or amplified by digital tools and directed at people because of their gender — has been documented as disproportionately affecting women in public life in India, though the country publishes no regular data on prosecutions.

    No individual has been charged. Darshan Thoogudeepa is not accused of any wrongdoing in the complaint, which concerns accounts said to be operated by his supporters.

  • Pakistan cyber agency announces seven arrests in three days

    Pakistan cyber agency announces seven arrests in three days

    Pakistan’s National Cyber Crime Investigation Agency announced the arrest of seven men in three days across Punjab province in cases involving the online harassment and blackmail of women.

    The agency said three men were detained on 28 July 2026 in operations in Lahore, Faisalabad and Multan, and a further four on 30 July in the same three cities.

    In the 28 July arrests, the agency named one man in Lahore, who it said had created a TikTok account to post a woman’s photographs while impersonating a government official in order to intimidate and blackmail her. It named another man in Faisalabad, accused of blackmailing a complainant’s sister through social media and WhatsApp, and a third man in Multan, accused of sharing images of a complainant and harassing her and her family.

    None of the men has entered a plea. The accounts of what they are alleged to have done come from the agency. Neither they nor their legal representatives were immediately available for comment.

    The pace of arrests is not unusual. The agency has announced detentions in similar cases repeatedly through July. What is not available is any figure for what happens afterwards. Pakistan has published no data on how many cyber-harassment arrests result in charges, how many charges reach trial, or how many trials end in conviction.

    Reporting on the same day pointed to the other end of the problem. JournalismPakistan’s Global Media Brief of 31 July found women journalists in the country continuing to face online abuse and AI-enabled disinformation alongside workplace discrimination and security threats.

  • EU deepfake labelling rules become enforceable

    EU deepfake labelling rules become enforceable

    Rules requiring artificial intelligence content to be labelled became enforceable across the European Union on 2 August 2026, with the European Commission’s AI Office and national authorities taking on enforcement.

    Under the transparency provisions of the AI Act, deepfakes — images, video or audio generated or edited using AI — must be labelled. AI-generated or altered content must also carry machine-readable marks so that it can be detected automatically. Chatbots and other interactive systems must tell users they are not dealing with a human.

    The obligation applies regardless of intent. Content that appears to depict a real person must be labelled even where there was no attempt to deceive, and, according to the Commission’s guidance, even where no real individual is depicted.

    Around 190 companies and organisations had signed the accompanying Code of Practice on transparency of AI-generated content by the end of July. Signatories can rely on its measures to demonstrate compliance with the labelling and detection requirements.

    The reach of the rules over technology-facilitated gender-based violence — abuse committed, assisted or amplified by digital tools and directed at people because of their gender — is untested. The provisions address disclosure rather than creation or distribution, and a labelled non-consensual sexual image remains a non-consensual sexual image.

    Specialists have questioned whether the rules can be enforced as drafted. Analysis published by Euronews on 28 July reported warnings that technical gaps and the absence of common standards for marking content could undermine enforcement in practice.

    The rules took effect two days after a federal judge in Minnesota allowed the first US state ban on nudification tools to come into force.

  • YouTube sent 1.8 million visits to nudify sites, research finds

    YouTube sent 1.8 million visits to nudify sites, research finds

    Mainstream social media platforms referred more than 5.7 million visits to websites and applications that generate non-consensual intimate images between December 2025 and March 2026, according to research published by the Institute for Strategic Dialogue.

    YouTube alone accounted for 1.82 million of those visits, more than 30 per cent of all referrals identified, the London-based research organisation found.

    The finding shifts attention from the tools themselves to the platforms that carry traffic to them. Regulatory action so far has concentrated on the applications and the app stores that distribute them. San Francisco’s city attorney wrote to Apple and Google in July over 13 named applications. Australia’s eSafety Commissioner has issued compliance directions to individual services.

    The Institute for Strategic Dialogue’s report, *Mapping the “Nudify” Tools Ecosystem*, is the outcome of a six-month project supported by the UK AI Safety Institute’s Challenge Fund. It traces how gendered deepfake content is created, circulated and sustained, and situates it within technology-facilitated gender-based violence, or TFGBV — abuse committed, assisted or amplified by digital tools and directed at people because of their gender.

    The report calls for what it describes as coordinated, system-wide responses spanning online, offline, hybrid and policy interventions, including increased platform regulation and funding for digital literacy workshops in schools.

    The research arrives as the European Union prepares to begin enforcing rules requiring AI-generated and altered content to be labelled, from 2 August. Those rules address disclosure rather than distribution, and would not on their face reach referral traffic of the kind the report describes.