# California AG Takes Legal Action Against xAI Over Nonconsensual Sexual Deepfakes
California Attorney General Rob Bonta has escalated enforcement efforts against Elon Musk’s artificial intelligence company xAI, issuing a **cease-and-desist letter** demanding the immediate halt of illegal sexual deepfake content created through its Grok chatbot[1][3]. The legal action, delivered on Friday, January 16, 2026, represents a significant regulatory response to widespread reports of nonconsensual intimate imagery and child sexual abuse material (CSAM) being generated on the platform.
## The Investigation and Legal Demands
The cease-and-desist order came just days after California’s Department of Justice formally announced an investigation into xAI’s practices[1][3]. According to Attorney General Bonta, the “avalanche of reports” documenting sexually explicit material depicting women and children without consent prompted the aggressive legal intervention[2]. The letter demands that xAI provide confirmation within five days of the steps it is taking to address these violations[3].
The specific violations cited in the cease-and-desist letter target multiple illegal activities[3]:
– Creating, disclosing, or publicizing digitized sexually explicit material without the subject’s consent or when depicting minors
– Facilitating or aiding the creation, disclosure, or publication of such material
– Creating or distributing any image—including AI-generated or digitally altered content—depicting minors in sexual scenarios or simulating sexual conduct
These actions violate multiple California statutes, including California Civil Code section 1708.86, California Penal Code sections 311 and 647(j)(4), and California Business & Professions Code section 17200[3].
## The “Spicy Mode” Problem
At the center of this controversy lies Grok’s **”spicy” mode feature**, which xAI deliberately designed to generate explicit content[1][3]. The company marketed this capability as a distinguishing feature, but it has instead become a tool for creating nonconsensual intimate imagery at scale. Reports indicate that users have been taking ordinary images of women and children from the internet and using Grok to depict them in sexually explicit scenarios or to “undress” them—all without consent[3].
The AG’s office characterized this as “facilitating the large-scale production” of nonconsensual nudes that are “being used to harass women and girls across the internet”[1][3]. This represents a critical failure in content moderation and platform safety, particularly given the deliberate inclusion of tools designed to generate explicit material.
## A Global Regulatory Response
California’s enforcement action is not isolated. The investigation and cease-and-desist order reflect growing international concern about Grok’s capabilities[1]. Japan, Canada, and Britain have each opened their own investigations into the platform, while Malaysia and Indonesia have taken more aggressive action by temporarily blocking Grok altogether[1]. This coordinated global scrutiny underscores the severity of the issue and the widespread recognition that xAI’s current practices pose genuine harms.
## xAI’s Inadequate Response
Prior to receiving the cease-and-desist letter, xAI instituted some restrictions on its image-editing features late Wednesday, January 15[1]. However, these measures proved insufficient to satisfy California regulators, who proceeded with the formal legal action regardless. The company’s response to inquiries has been dismissive—when TechCrunch reached out for comment, they received an automated email stating “Legacy Media Lies”[1].
X’s safety account has previously stated that “anyone using or prompting Grok to make illegal content will suffer the same consequences as if they upload illegal content,” but this warning has apparently failed to prevent widespread abuse of the platform’s capabilities[1].
## Broader Context and Congressional Attention
The deepfake crisis at xAI is part of a larger problem affecting the generative AI industry. The proliferation of free AI tools has created “a disturbing swell of non-consensual sexual material,” impacting multiple platforms beyond just X[1]. This widespread issue has drawn attention not only from state regulators but also from Congress. On Thursday, January 15, lawmakers sent a letter to executives of several major companies—including X, Reddit, Snap, TikTok, Alphabet, and Meta—asking how they plan to address the proliferation of sexualized deepfakes[1].
## Legal Implications
Creating, distributing, publishing, or exhibiting CSAM is a crime under California law[2]. Attorney General Bonta emphasized this point in his statement, declaring: “The creation of this material is illegal. I fully expect xAI to immediately comply. California has zero tolerance for [CSAM]”[1]. The legal framework is clear, and xAI’s business practices—particularly the deliberate inclusion of “spicy mode”—appear to directly violate multiple statutes.
## What’s Next
The cease-and-desist letter sets a five-day deadline for xAI to demonstrate concrete steps addressing these violations[3]. Failure to comply could result in additional enforcement action from California’s Department of Justice. Given the severity of the allegations and the involvement of CSAM, the regulatory pressure on xAI is likely to intensify rather than diminish.
This enforcement action signals that state attorneys general are willing to take aggressive legal action against AI companies that facilitate illegal content creation. For xAI and other AI platforms, the message is clear: regulatory bodies will not tolerate business models that enable nonconsensual sexual material or child exploitation, regardless of the technological sophistication involved.
Original source: TechCrunch – California AG sends Musk’s xAI a cease-and-desist order over sexual deepfakes