One in six congresswomen reportedly targeted by AI-generated nonconsensual intimate imagery
Research by the American Sunlight Project found that approximately one in six congresswomen have been targeted by AI-generated nonconsensual intimate imagery on websites hosting deepfake pornography. Women lawmakers were 70 times more likely than male counterparts to be depicted, representing a stark disparity in the gendered abuse of synthetic media technology.
What happened
The American Sunlight Project analyzed 11 websites known to host sexual deepfake content and found 35,239 mentions of U.S. lawmakers. The research identified 26 members of Congress and senators whose names appeared on these sites, including 25 women and one man.
Women members of Congress were 70 times more likely than men to be depicted in deepfake images, with approximately one in six congresswomen reportedly targeted by nonconsensual intimate imagery.
Reported impact
The gendered disparity in targeting reflects a precarious online environment for women leaders navigating sexual abuse risks through AI-generated imagery.
Women lawmakers had deepfakes removed within 48 hours following reports, according to the American Sunlight Project findings.
Index analysis
The research found no link between political party or state representation and deepfake targeting, though older members of Congress were slightly less likely to appear in such imagery.
Governance controls
In January, a bipartisan group of senators introduced the Disrupt Explicit Forged Images and Non-Consensual Edits Act of 2024 (Defiance Act), intended to criminalize spreading nonconsensual sexualized digital forgeries made with AI. The bill passed the Senate in July but remained under House consideration.
The Senate passed the Take It Down Act, which criminalizes publishing digitally manipulated deepfakes online and imposes penalties on companies failing to remove such content within 48 hours of being reported.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.