Poll finds broad alarm over AI sexual deepfakes and election disinformation

A new nationwide poll from Women Who Tech and Rad Campaign finds Americans want stronger limits on AI used to sexualize women and children and to spread election disinformation. The survey also shows widespread skepticism about Big Tech’s handling of misinformation, with concern highest around deepfakes tied to the 2026 midterms.

Why it matters: - The poll fills a data gap on public concern over AI-generated sexual imagery targeting women and children. - The findings arrive as the 2026 midterm campaign intensifies and AI deepfakes are spreading faster online. - The results suggest broad public support for federal guardrails on AI in elections and on platforms. - Congress has not enacted federal election deepfake rules, leaving states to step in on a patchwork basis. - Two AI sexual imagery bills, the DEFIANCE Act and the ENFORCE Act, have passed the Senate unanimously and are now in the House.

What happened: - Women Who Tech and Rad Campaign released a nationwide poll published at AIpolldata.com. - The survey was fielded June 13-18, 2026. - The poll measured concerns about AI sexualization, election misinformation, political engagement, personal AI use, and views on AI’s impact on jobs and the economy. - The survey followed the Grok sexual deepfake scandal on X, where users posted AI-generated sexual imagery, mostly of women and children.

The details: - 71% of respondents said they have some or high concern about AI deepfakes that sexualize children. - 67% said they have some or high concern about AI being used to create sexual imagery of women. - 49% said they are very concerned about AI spreading disinformation ahead of the November midterm elections. - 73% said they have at least some concern about election-related AI disinformation. - 81% favor either requiring disclosure of AI-generated content or banning its use on official government accounts. - 42% support mandatory disclosure of AI-generated content. - 39% support prohibiting AI-generated content on official government accounts. - 19% think politicians should be free to use AI-generated content on social media without restriction. - About half of social media users say misinformation is more common on major platforms than before. - 51% of X users say misinformation or disinformation is “more” or “a lot more” common. - 48% say the same about Instagram. - 48% say the same about TikTok. - 47% say the same about Facebook. - 93% of Instagram users say they see the same amount or more politics-related misinformation and disinformation than a few years ago. - 91% say that about Facebook. - 90% say that about YouTube. - 88% say that about X. - 87% say that about TikTok. - College-educated respondents show higher concern about AI misuse in politics, at 50%. - Registered voters show higher concern at 47%. - Non-college-educated respondents show lower concern at 41%. - Non-registered respondents show lower concern at 35%. - The poll also included demographic breakouts for optimism about AI and how often people use AI personally.

Between the lines: - The survey points to a public that is more wary of AI harms than excited about AI benefits. - The strongest alarm centers on sexual exploitation and election manipulation, two areas where regulation is politically easier to justify. - The mixed federal response leaves companies, states, and voters stuck with uneven protections. - Allyson Kapin, founder of Women Who Tech and Rad Campaign, said policymakers and Big Tech are lagging behind constituents and customers.

What's next: - The House faces pressure to act on the DEFIANCE Act and ENFORCE Act after Senate passage. - Lawmakers are also likely to face more calls for rules on AI-generated political content during the 2026 campaign. - State governments are expected to keep filling gaps as federal action stalls. - The poll’s full results, visualizations and methodology are available through the published data page.