.avif){kind=small}
Activist and anti-image-based abuse advocate [PERSON_001] was targeted by a deepfake attack where her likeness was superimposed onto pornographic content. This incident, which followed her advocacy work against unverified content on PornHub, resulted in significant online abuse, doxing, and threats of violence against her.
Kate Isaacs, a London-based activist and founder of the #NotYourPorn campaign, was targeted in a deepfake incident. Her face was alleged to have been digitally manipulated onto a pornographic video using AI and shared online. The reported video, tagged with her name, is alleged to have led to streams of abuse, doxing, and threats of violence. The attack reportedly followed her efforts to pressure PornHub to remove unverified content.
The incident involves the targeted creation of humiliating and sexual deepfake imagery of a specific victim to silence and harass her, which fits the boundary rule for targeted humiliation under 4.3.
The harm was caused by human actors intentionally using deployed AI deepfake tools to target, humiliate, and harass the victims.
Limited Risk: The report describes AI-generated content such as deepfakes, which fall under Risk Level 3 and are subject to specific transparency obligations under the EU AI Act.
Highest direct severity in any category: Severe. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.
Reported: The report explicitly describes the creation and distribution of toxic and malicious content, specifically non-consensual deepfake pornography.
Directly caused: AI deepfake software was used to generate sexually explicit videos and images of the victims without their consent, which were shared on public platforms like Twitter/X.
Indirectly caused: The posting of the deepfakes led to streams of abusive comments, rape threats, and doxing on social media.
Inferred additional harm: Millions of other deepfake pornographic videos are hosted on specialized websites, exposing millions of users to non-consensual toxic content.
Reported: The report explicitly describes differential treatment, highlighting that this is a highly gendered crime primarily targeting women.
Directly caused: Women's public images were disproportionately harvested and weaponized to create sexually explicit content, subjecting them to systemic harassment and humiliation compared to men.
Indirectly caused: The societal shaming and harassment of female victims, such as the politician being propositioned on the street based on the fake video.
Inferred additional harm: Widespread systemic discrimination and harassment of women online through the proliferation of easy-to-use deepfake apps.
Reported: The report explicitly describes violations of human and civil rights, including bodily autonomy, dignity, and freedom from harassment.
Directly caused: The non-consensual use of the victims' identities and likenesses in sexual content violated their fundamental right to personal dignity and privacy.
Indirectly caused: The threats and harassment aimed to silence the campaigner [PERSON_001] and the politician, infringing on their freedom of expression and political participation.
Inferred additional harm: Thousands of women having their civil rights and personal safety compromised by non-consensual synthetic pornography.
Reported: The report explicitly describes an attack on a political candidate during an active election campaign.
Directly caused: N/A
Indirectly caused: A Northern Irish politician was targeted with fake porn during her election campaign in April 2022 to discredit and silence her, undermining democratic participation.
Inferred additional harm: The proliferation of deepfakes targeting female politicians could discourage women from running for public office, eroding democratic representation.
Reported: The report describes privacy violations, but these were carried out by human actors rather than the AI system itself.
Directly caused: N/A
Indirectly caused: Human perpetrators doxed [PERSON_001] by posting her work and home addresses online alongside the deepfake video.
Inferred additional harm: N/A
Reported: The report explicitly describes severe psychological and mental-health harm to multiple victims, including panic, terror, anxiety, paranoia, and trauma.
Directly caused: [PERSON_001] experienced severe panic, terror, and paranoia, causing her to restrict her movements and delete her Twitter account. Other victims reported feeling devastated, violated, and living in a constant state of fear.
Indirectly caused: The subsequent doxing, rape threats, and street harassment resulting from the deepfakes caused additional severe trauma and fear for personal safety.
Inferred additional harm: It is highly likely that thousands of other unquantified victims of these deepfake platforms suffer similar severe psychological trauma, anxiety, and depression.
Reported: The report explicitly describes epistemic harm, where highly convincing fake videos led people to believe the victims actually performed in pornographic videos.
Directly caused: AI software generated highly convincing visual hoaxes of the victims.
Indirectly caused: Members of the public believed the fake videos were real, leading to street harassment and public jeering of the politician.
Inferred additional harm: Widespread erosion of trust in digital media and video evidence due to the ubiquity of convincing deepfakes.
This incident involves the targeted creation of non-consensual deepfake pornography against an activist and a political candidate using consumer-grade AI tools. While causing severe psychological harm, doxing, and localized democratic disruption to a political campaign, the national security threat remains minor and manageable within standard law enforcement frameworks.