A new nationwide poll finds broad, bipartisan alarm over AI-generated sexual imagery of women and children — and it lands at the exact moment two bills that could address it are stuck in a Republican-controlled House.
The survey, conducted June 13–18 by Women Who Tech and Rad Campaign, is the first to measure public concern about AI sexualization of women, and the first in two years to focus specifically on AI sexualization of children. The findings are stark: 71% of respondents register some or high concern about AI deepfakes that sexualize kids, and 67% feel the same about AI-generated sexual imagery of women.
Those numbers aren’t coming from nowhere. The poll follows a recent scandal on Twitter, where the platform’s built-in AI image tool was used to flood the site with sexually explicit deepfakes, most of them of women and children — a wave so severe that Apple reportedly threatened to pull the app from its App Store.
The bills sitting in the House
Two pieces of legislation directly address the concerns expressed in the poll. The DEFIANCE Act — sometimes called the Paris Hilton bill — would let victims of nonconsensual sexual deepfakes sue the people who made and spread them, with damages up to $250,000. The ENFORCE Act would let prosecutors treat AI-generated child sexual abuse material the same as any other child sexual abuse material.
Both passed the Senate unanimously, basically unheard of these days. And yet both are stalled, still sitting in the Republican-controlled House.
The DEFIANCE Act has bipartisan support — it has nearly 60 bipartisan co-sponsors in the House, and Speaker Mike Johnson himself claims he supports it.
What it’s missing, however, is a floor vote, which only Johnson can schedule.
AOC, the bill’s lead House sponsor, thinks she knows why: “There’s been some reporting recently that some of the block around this may literally just be because my name is attached to it,” she told The Independent.
“I don’t think that we should be emboldening child pornographers and helping abusers just because of some personal beef that they may have.”
ENFORCE has no such explanation. It passed the Senate back in December, over seven months ago, and nobody in the House has said why it hasn’t moved. It’s simply not been a priority.
And some companies seem to be actively trying to make it worse. Elon Musk’s xAI is currently suing Minnesota to block a state law banning tools that use AI to generate fake nude images of real people without their consent.
“The poll clearly shows that policymakers and Big Tech companies are lagging far behind their own constituents and customers on this issue,” said Allyson Kapin, founder of Women Who Tech and founding partner at Rad Campaign.
Voters “may disagree about whether AI will do anything good for them personally, but there’s a clear bipartisan majority for safeguards against sexual deepfakes of women and children and for guardrails on politicians’ use of AI.”
It’s not just deepfakes — it’s elections, too
The same poll found deep unease about AI’s role in politics more broadly.
Nearly half of respondents, Democrats and Republicans alike, said they’re “very concerned” about AI-generated disinformation ahead of the November midterms, and 73% expressed some level of concern. Eighty-one percent want AI-generated political content either disclosed or banned outright on official government accounts.
Only 19% think politicians should be able to use AI content with no restrictions at all — and that 19% isn’t evenly spread. Republicans are more than twice as likely as Democrats to say politicians should face no restrictions at all.
Which is telling, as Republicans are the ones actually flooding races with this stuff. There’s no federal law banning it.
In Kentucky’s GOP primary, PAC-funded AI deepfakes helped sink longtime Rep. Thomas Massie — fabricated footage placed him in a hotel scene with Reps. Alexandria Ocasio-Cortez and Ilhan Omar.
In June, Texas Governor Greg Abbott posted a transphobic AI-generated image depicting Texas Democratic Senate candidate James Talarico dressed as a woman. A pro-Trump PAC spent six figures on its own transphobic AI ad of Talarico.
The response so far: not much
Federal action has been thin. The TAKE IT DOWN Act, which took effect in May, requires platforms to remove non-consensual sexual imagery — including deepfakes — within 48 hours of a request. Supporters say it’s a start, but it’s reactive — it only removes content after someone’s already been victimized, and it does nothing to stop the people making it in the first place.
Meanwhile, Trump has pushed to strip states of their power to regulate AI at all.
Republicans could have a genuine bipartisan win here. Instead, they’re leaving women and children exposed so AOC doesn’t get a perceived win.
The poll surveyed 1,015 adults and has a margin of error of roughly ±3.1 percentage points at the 95% confidence level. Full results and methodology are available at radcampaign.com.