A trio of US Senators say Apple and Google are failing to effectively detect and remove apps that can be used to generate nonconsensual nude or pornographic imagery.
asserted
that → say → imagery
They also say Meta is continuing to profit off of ads for these so-called nudification apps that “victimize women and girls worldwide”.
asserted
that → say → women
On Wednesday, senators Ron Wyden, Ed Markey and Ben Luján sent letters to the CEOs of the tech companies, saying the safeguards they have in place to keep nudification off their app stores or platforms are not sufficient.
asserted
they → send → platforms
The senators sent the letters after multiple studies by the Tech Transparency Project (TTP), the latest of which was exclusively reported in the Guardian, found that these apps are either available on Apple and Google app stores or are being advertised in ads on Meta platforms in spite of company policies that prohibit them.
asserted
that → send → them
According to the studies, children continue to have easy access to these apps.
uncertain
children → accord → apps
Senator Wyden of Oregon said Congress will need to act if the companies “don’t do a better job of cracking down on apps expressly designed and marketed to create harm”.
asserted
companies → say → harm
“AI has supercharged the ability to create and share nonconsensual sexual images, creating tremendous harm to women and girls,” Wyden said in a statement to the Guardian US.
asserted
Wyden → supercharge → US
“Instead of ramping up investment to meet this challenge, big tech companies are cutting back on the very teams responsible for protecting users from harm.
asserted
companies → ramp → harm
I’m looking for answers about why Meta, Apple and Google are missing obvious violations of their own rules.”
asserted
Meta → look → rules
In a statement, Karl Ryan, a Google spokesperson, said: “Google Play does not allow apps that contain sexual content or allow users to nudify images of people, and we continually take proactive steps to detect and remove apps with harmful content.”
asserted
we → say → content
A Meta spokesman, Andy Stone, said in a statement: “We prohibit CSAM and the promotion of nudify apps, and we work aggressively to keep this material off our services.
asserted
we → say → services
The criminals behind it take increasingly deceptive steps to evade detection – hiding behind fake names and identities, running ads that look benign, and sometimes linking to Google’s and Apple’s app stores, with the ads’ true intent revealed only after clicking through.
asserted
intent → take → stores
We’ve taken extensive steps against these perpetrators, from banning Keling Technology Limited and taking legal action against other advertisers, to going after the coordinated networks and removing 92 million sexual ads – and we’re actively exploring new ways to improve our detection.
asserted
we → take → detection
The senators wrote that AI-generated nonconsensual nude imagery has seen an uptick in the last two years.
asserted
imagery → write → years
Surveys bear that assertion out.
asserted
Surveys → bear → assertion
Unicef reported at the start of 2026 that at least 1.2 million children across 11 countries reported being victimized of nonconsensual nude images.
asserted
children → report → images
Law enforcement have also warned in various intelligence memos that officers should remain vigilant because AI may make it easier to circumvent CSAM (child sexual abuse material) detection methods, according to documents the Guardian previously reported on.
uncertain
Guardian → warn → documents
Apple clarified its policies after TTP researchers reached out about nudification apps they found on the app store in October.
asserted
they → clarify → October
Google previously told the Guardian that the company proactively works to detect and remove apps with harmful content and Meta introduced new AI tools to detect and remove ads that contain CSAM.
asserted
that → tell → CSAM
Keith Chu, the chief communications adviser and deputy policy director to Wyden, said that clarifying policies is not the same thing as enforcing them.
asserted
clarifying → say → them
Despite the companies reaffirming the prohibition of these apps, they continue to proliferate across Apple and Google app stores or be promoted in ads on Meta platforms, he added.
asserted
he → reaffirm → platforms
In the latest TTP study in October, which the Guardian first reported, researchers found that they were able to access dozens of apps with features that enabled the creation of AI-generated nude and pornographic imagery despite indicating the accounts they created were for 13-year-old boys.
asserted
they → report → boys
In response to inquiries from TTP and the Guardian US, Apple and Google removed the apps that were mentioned in the studies.
asserted
that → remove → studies
However, the apps remained functional on phones they were already downloaded on.
asserted
they → remain → phones
These findings suggest that Apple and Google are “not merely failing to remove applications that facilitate intimate image abuse, it may also be profiting from their distribution”, the letter from the senators read.
uncertain
letter → suggest → senators
Katie Paul, the director of TTP, said that though Apple and Google say they review every app on their platform the companies do not disclose what that review entails.
asserted
review → say → what
“Vague descriptions of how each company defines ‘nonconsensual intimate imagery’ and lack of detail in how both companies actually go about app review allow them to have it both ways: claiming they have policies on the books, and applying those policies based on what best fits the narrative at any given time,” Paul said.
asserted
Paul → define → time
Apple, for its part, also said that some developers will pull a “bait-and-switch” and add features that enable users to create AI-generated nude or pornographic imagery after the iPhone maker approves an app.
asserted
maker → say → app
Wyden’s office said that they expect the company to review apps on an ongoing basis.
asserted
company → say → basis
On Meta’s social networks, however, many of these developers advertise much more explicitly the nudification functions their apps include.
asserted
apps → advertise → functions
TTP and other research groups have discovered hundreds of paid ads on Instagram and Facebook that include AI-generated sexualized images of real children and other CSAM.
asserted
that → discover → children
In September 2026, TTP found 332 paid ads on Facebook and Instagram that included CSAM.
asserted
that → find → CSAM
This month, AIForensics, a European non-profit research organization, found that some apps that were hiding their AI nudification features from Apple and Google were explicitly promoting “sexual deepfake functionality” in ads on Meta platforms.
asserted
that → find → platforms
“Despite Meta’s stated prohibition on these ads, independent investigations have documented extensive ad campaigns on Facebook and Instagram directing customers to apps capable of digitally undressing people in photographs or placing their likeness into sexually explicit content without their consent,” the senators wrote in their letter to Meta’s CEO, Mark Zuckerberg.
asserted
senators → state → CEO
In the letter, the senators asked how ads containing apparent CSAM passed the company’s review systems.
asserted
ads → ask → systems
On 7 October, Meta said that it had taken additional measures to combat ads with CSAM, however, Paul said TTP researchers have reported dozens of ads containing CSAM to the National Center for Missing & Exploited Children since then.
asserted
researchers → say → Children
Screenshots TTP shared with the Guardian show images of an ad for an app to create “AI sex friends” highlighting images of young girls.
asserted
TTP → share → girls
These ads were active for several hours before being taken down but remain available in Meta’s ad library.
asserted
ads → take → library
The organization has “yet to identify ads featuring videos of CSAM run by any other ad platforms”.
asserted
organization → identify → platforms
“Meta has repeatedly boasted about its sophisticated AI systems for reviewing harms on platforms, but for more than a year those systems have repeatedly failed to flag ads featuring graphic sexual videos of children as young as toddlers,” Paul said.
asserted
Paul → boast → toddlers