SACRAMENTO, Calif. — California announced an investigation Wednesday into Elon Musk's xAI, with Gov. Gavin Newsom accusing the artificial intelligence company of becoming a "breeding ground for predators to spread nonconsensual sexually explicit AI deepfakes."

Grok, the xAI chatbot, includes image-generation features that allow users to morph existing photos into new images. The newly created images are then posted publicly on X.

Critics say there are not sufficient safeguards on what images can be generated, prompting an influx of sexually explicit or nonconsensual images based on real people, including altered depictions that appear to show individuals partially or fully undressed. Others have generated images that appear to show minors, prompting criticism that the tools are being used to create child pornography.

The social media site has previously said, "we take action against illegal content on X, including Child Sexual Abuse Material (CSAM), by removing it, permanently suspending accounts, and working with local governments and law enforcement as necessary."

Earlier this month, Grok started limiting the ability for nonpaying users to create sexualized images amid a global outcry from users and governments.

Newsom called the images being created "vile." Atty. Gen. Rob Bonta said his office will use "all tools at our disposal to keep Californians safe."

"The avalanche of reports detailing the non-consensual, sexually explicit material that xAI has produced and posted online in recent weeks is shocking," Bonta said in a statement Wednesday. "This material, which depicts women and children in nude and sexually explicit situations, has been used to harass people across the internet."

Newsom signed a pair of bills in 2024 that made it illegal to create, possess or distribute sexually charged images of minors even when they're created with computers, not cameras. The measures took effect last year.

Assembly Bill 1831, authored by Assemblymember Marc Berman, D-Menlo Park, expanded the state's child porn prohibition to material that "contains a digitally altered or artificial-intelligence-generated depiction (of) what appears to be a person under 18 years of age" engaging in or simulating sexual conduct.

Senate Bill 1381, authored by Sen. Aisha Wahab, D-Hayward, amended state law to more clearly prohibit using AI to create images of real children engaged in sexual conduct, or using children as models for digitally altered or AI-generated child pornography.

Last year, Newsom signed two more bills into law that make it clear that people and companies — not AI tools — are legally responsible for harm, strengthening victims' ability to sue over nonconsensual deepfake pornography and preventing AI developers or users from claiming the technology acted on its own.

While announcing the state investigation Wednesday, Bonta also urged xAI to take immediate action to ensure it is following California law. His office said potential victims can file a complaint at oag.ca.gov/report.

"We have zero tolerance for the AI-based creation and dissemination of nonconsensual intimate images or of child sexual abuse material," Bonta said.

One researcher whose analysis was published by Bloomberg found that X has become the most prolific platform for digitally undressing people. Genevieve Oh, a researcher who studies social media and deepfakes, found that during a 24-hour review, the @Grok account generated about 6,700 images per hour that were sexually suggestive or appeared to digitally undress people — compared with an average of 79 new images per hour across the five other leading deepfake sites.

One prominent trend using Grok involves users asking the chatbot to create pictures of celebrities, politicians and private individuals in bikinis. The targets of those prompts are often women and girls.

On Wednesday, 28 organizations that included women's groups, tech watchdogs and progressive activists published an open letter calling on both Google and Apple to remove X and Grok from their app stores because of the sexualized material.

"The proliferation of non-consensual deepfakes has irreversibly altered the lives of women and children who've been completely stripped of their privacy, autonomy, and safety," Jenna Sherman, campaign director at the gender justice organization UltraViolet, said in a statement. "This may be a game to Elon Musk — who instead of taking down Grok entirely until proper safeguards were implemented instead made AI image abuse a paid feature — but it's deadly serious for the thousands of users actively being harmed."

The European Commission opened inquiries into Grok's algorithms and potential dissemination of illegal content, ordering Musk's company to preserve documents related to the chatbot's development.

Sweden joined the criticism of how Grok is being used to target women after Deputy Prime Minister Ebba Busch became a target of users who generated a picture of her in a bikini. Busch said the picture was "actually a pretty good picture."

"The problem is I was involuntarily undressed by Elon Musk's Grok on X," Busch said in a video posted on X. "Let me be very clear, as a woman I decide when, where and to whom I show myself in a bikini."

In the video Busch uploaded, she's dressed in heavy winter gear, saying she was hunting in the snow. Several users responded to the video with newly generated Grok images and videos showing Busch in a bikini.

_____

©2026 Los Angeles Times. Visit latimes.com. Distributed by Tribune Content Agency, LLC.