Controversy surrounding Elon Musk's AI chatbot Grok, stemming from its generation of non-consensual explicit imagery, continues to escalate, with US law enforcement now formally involved. On Wednesday, Eastern Time, California Attorney General Rob Bonta announced the launch of an investigation into xAI and its AI model Grok, marking a significant escalation of the issue from global public condemnation to concrete legal action.
According to an announcement on the California Department of Justice's official website, the investigation focuses on Grok's "mass production" of non-consensual deepfake intimate images, which are allegedly being used to harass women and girls online. The probe will determine if and how xAI violated the law. California Governor Gavin Newsom has also voiced support for the investigation, labeling xAI's actions "disgusting." This represents the latest regulatory challenge for Grok on a global scale. Previously, governments including the EU, UK, France, Malaysia, Indonesia, India, and Brazil have already taken action or issued warnings concerning the tool. In response, Musk stated this week that he was unaware of the specific situation and emphasized that Grok is programmed to reject illegal requests. This development coincides with the US Department of Defense's announcement to deploy Grok on both classified and unclassified networks, highlighting the urgent need for AI governance as the tension between regulation and application grows.
California Launches Formal Investigation, Accusing xAI of Facilitating Image Abuse
In an official statement on Wednesday, California Attorney General Rob Bonta stated, "The numerous reports over the past few weeks of xAI producing and publishing non-consensual explicit material online are shocking." He stressed that these materials, depicting women and children in nude and explicit sexual scenarios, have been used to harass victims across the internet. Per the Department of Justice announcement, the investigation targets xAI for "appear[ing] to be facilitating the mass production of deepfake non-consensual intimate images, which are being used to harass women and girls online, including via the social media platform X." Bonta explicitly stated in the declaration:
"We have zero tolerance for the generation or dissemination of AI-based non-consensual intimate imagery or child sexual abuse material."
The announcement detailed the issues with Grok. It mentioned that over recent weeks, numerous news reports indicated that Grok users took ordinary photos of women and children from the internet and used Grok to depict these individuals in suggestive and explicit sexual scenarios, even "stripping" their clothes, all without the subjects' knowledge or consent. The announcement stated:
"Even more alarming, news reports indicate that individuals have used Grok to alter photos of children, depicting them in revealing clothing and in sexual scenarios. It has also been reported that xAI generated realistic images of children engaged in sexual activity. These uses of Grok appear to be happening at scale. According to one analysis, of 20,000 images generated by xAI between Christmas and New Year's, over half depicted individuals in revealing attire, some of whom appeared to be children."
xAI's "Spicy Mode" as a Marketing Feature
The California DOJ's investigation focuses on xAI's product design choices. The announcement pointed out that xAI developed Grok's image generation model to include what the company calls a "spicy mode," which generates explicit content. "The company has marketed this mode as a selling point, which has unsurprisingly led to a surge in non-consensual sexualized content featuring individuals." The problem emerged following the launch of Grok Imagine last year. This AI image generator allows users to create videos and pictures from text prompts, including the "spicy mode" capable of generating adult content. The situation deteriorated sharply late last month when Grok, hosted on the X platform, apparently began approving a large volume of user requests to modify images posted by others, with requests such as "make her wear a transparent bikini." Musk has positioned his chatbot as a less restricted alternative with fewer safeguards than its competitors. Combined with the fact that images generated by Grok are publicly visible and thus easily shareable, this has amplified the severity of the problem.
Restrictive Measures Prove Ineffective
In response to the controversy, Grok indicated it would restrict image generation and editing features to paying subscribers. Last Friday, the chatbot responded to image modification requests with: "Image generation and editing are currently limited to paid subscribers. You can subscribe to unlock these features." It reportedly granted public request visibility only to X users with a blue verification checkmark, who pay $8 monthly for premium features. However, as confirmed on Tuesday, free users could still access the image editing tool by clicking the "edit image" button displayed on each picture on the platform, rather than by tagging Grok with a request. Independent Grok websites and applications were still approving image edit requests on Tuesday. In response to requests for comment, Musk's xAI company has consistently replied with an automated message stating "traditional media lies." X has previously stated it would take action to remove illegal content, including child sexual abuse material, permanently suspend user accounts, "and cooperate with local government and law enforcement where required." Musk has warned that "anyone using Grok to produce illegal content will face the same consequences as someone uploading illegal content."
Global Regulatory Pressure Intensifies
Prior to the California investigation, Grok had already faced regulatory actions in multiple jurisdictions worldwide. UK media regulator Ofcom launched an investigation on Monday, examining whether Grok breached its duties to protect people from illegal content. Indonesia temporarily blocked access to Grok last Saturday, and the Malaysian government stated on Tuesday that it would take legal action against X and xAI. The EU's top technology official warned that X must address the issue, citing the Digital Services Act and stating, "failing this, we will not hesitate to fully use the DSA to protect EU citizens." The Paris Prosecutor's Office in France stated it is expanding its investigation into X to include explicit sexual deepfake content. According to a report, Musk responded this Wednesday by stating he was unaware of any situation regarding Grok generating "nude images of minors," emphasizing that Grok's programming is set to reject illegal requests and that it must comply with the laws of all countries. Currently, multiple governments and organizations advocating for women's and children's rights have called on Apple and Google to remove the application from their respective app stores.
Comments