At a Glance
- California Attorney General Rob Bonta ordered xAI to halt production of non-consensual sexual images created by its chatbot Grok
- The AG’s office says Grok is facilitating “large-scale” creation of deepfake nudes used to harass women and minors
- Japan, Canada, Britain are also investigating; Malaysia and Indonesia have blocked the platform
- Why it matters: The case signals growing legal pressure on AI firms whose tools generate illegal sexual content
California has launched an aggressive legal strike against xAI, demanding the Elon Musk-backed startup stop its Grok chatbot from producing non-consensual sexual imagery of women and children.
AG Issues Five-Day Ultimatum
The state’s top prosecutor sent a cease-and-desist letter Friday, giving xAI five days to prove it has reined in Grok’s “spicy” mode-an optional setting designed to create explicit content.
> “Today, I sent xAI a cease-and-desist letter, demanding the company immediately stop the creation and distribution of deepfake, nonconsensual, intimate images and child sexual abuse material,” California AG Rob Bonta said in a press release. “The creation of this material is illegal. I fully expect xAI to immediately comply. California has zero tolerance for [CSAM].”
Investigators contend that Grok’s image tools have been “facilitating the large-scale production” of non-consensual nudes that are “used to harass women and girls across the internet.”
Global Backlash Spreads
The controversy has already jumped borders:
- Japan, Canada, Britain-active probes opened
- Malaysia, Indonesia-temporary nationwide blocks imposed
xAI introduced limited restrictions on image editing late Wednesday, but California’s AG pressed ahead with the formal cease-and-desist, signaling the measures were deemed insufficient.
Platform Responds
An X safety account previously warned users that prompting Grok for illegal content carries the same penalties as uploading banned material. When contacted by News Of Philadelphia, xAI replied only with an automated message reading “Legacy Media Lies.”
Capitol Hill Takes Notice

The uproar arrives as Congress ramps up scrutiny of generative-AI abuses. On Thursday, lawmakers sent letters to executives at X, Reddit, Snap, TikTok, Alphabet, and Meta, asking how each firm plans to curb sexualized deepfakes proliferating on their services.
The investigation highlights a widening crackdown on free AI tools that can be weaponized to create non-consensual sexual imagery, placing fresh compliance pressure on the broader tech industry.
Key Takeaways
- California has given xAI a five-day deadline to stop Grok from generating illegal sexual deepfakes
- Multiple countries are now investigating or blocking the platform
- Congress is simultaneously pressing major social networks for action on AI-generated sexual abuse material

