Analysis
San Francisco City Attorney David Chiu issued formal letters to Apple and Google on July 17 demanding both companies remove AI-powered 'nudify' apps -- tools that generate non-consensual deepfake nude images -- from their app stores, citing a 2025 California law that permits civil actions against third parties that knowingly facilitate such material, according to TechCrunch and Ars Technica.
Apple confirmed it has already removed three apps named in the letter and is investigating four others, and said it is 'in the process of terminating their developer accounts.' Google said all five apps cited in its letter have been suspended, adding that when violations are reported, the company investigates and has 'suspended hundreds of violating apps' historically. Both companies now face a 28-day deadline to fully respond to Chiu's demands.
The action follows two prior warnings from the Tech Transparency Project, which first flagged nudify apps on both platforms in January 2025 and issued a follow-up report in April 2025 documenting continued violations despite the initial warning -- meaning this week's city-attorney letters arrive after roughly 18 months of documented, unresolved policy failures at both companies.
“Both companies now face a 28-day deadline to fully respond to Chiu's demands.”
Chiu's most pointed claim is financial rather than purely regulatory: he told reporters Apple and Google have likely earned 'millions of dollars in fees' from apps that generate non-consensual intimate images of women and girls, framing the issue explicitly as a case of platforms profiting from harm rather than merely failing to catch it. The action targets the app-store distribution layer specifically, a different enforcement angle than the AI-model-developer liability debates that have dominated deepfake policy discussion so far.
The case sits alongside a broader wave of state-level AI-harm enforcement actions in 2026, as federal AI regulation remains largely stalled and state attorneys general and city law offices increasingly fill the gap with civil actions under existing consumer-protection and image-based-abuse statutes.
The bear case: removing named apps doesn't address the underlying open-source models and techniques that make nudify tools trivial to rebuild and redistribute under a new app name, and app-store enforcement is inherently reactive rather than preventive. What to watch next: whether Apple and Google's 28-day responses include proactive detection changes, or just the removal of the specifically named apps.
