<p class="wp-block-paragraph">A technology and artificial intelligence safety group has sued OpenAI over a July incident in which the company’s technologies autonomously gained access to the systems of AI developer hub Hugging Face.</p>
<p class="wp-block-paragraph">Legal Advocates for Safe Science and Technology (LASST) filed the lawsuit on Thursday in the Superior Court of California in San Francisco, arguing that the company’s conduct violated the state’s anti-hacking laws. LASST contends that OpenAI’s defense that “artificial intelligence autonomously caused the harm” in the Hugging Face incident is insufficient, and claims the company is “on the hook for harms” caused in the hacking.</p>
<p class="wp-block-paragraph">“OpenAI is both the developer and deployer of the AI that caused the harm, and it is thus responsible,” LASST wrote in the filing.</p>
<p class="wp-block-paragraph">Around 700 of OpenAI’s agents went rogue and autonomously hacked into Hugging Face in July, with the incident spanning a six-day period. OpenAI CEO Sam Altman stated it was the first security incident he felt “viscerally.” Since then, OpenAI has reported other incidents of agents acting without authorization from humans, raising concerns from AI safety advocates.</p>
<p class="wp-block-paragraph">LASST’s lawsuit is reportedly the first legal complaint against the July hack of the AI startup, according to Politico.</p>
<p class="wp-block-paragraph">The advocacy group points to California’s Comprehensive Computer Data Access and Fraud Act and argues that OpenAI violated the Unfair Competition Law. LASST also claims OpenAI caused damage to the group itself, stating it has had “to divert resources from its normal activities to educate regulators, civil society, and the public about the facts, legal issues, and potential dangers of OpenAI’s conduct relating to its hack of Hugging Face.”</p>
<p class="wp-block-paragraph">The lawsuit requests the court for an injunction that would prevent OpenAI or its agents from accessing computer networks or systems without authorization. It also seeks an injunction from the court on OpenAI engaging in “unlawful or unfair business practices” or in business practices that could cause public harm.</p>
<p class="wp-block-paragraph">The complaint arises as the AI industry addresses “increasingly cyber-capable models,” as OpenAI has described, and industry leaders call for a slowdown in frontier AI model development. Anthropic’s Dario Amodei was the first American AI company CEO to advocate for a slowdown and national guardrails, but Altman quickly supported this call.</p>
<p class="wp-block-paragraph"><strong><a href="https://www.washingtonexaminer.com/policy/technology/4746306/openai-scraps-release-astra-model-safety-concerns/">OPENAI SCRAPS RELEASE OF NEW ‘ASTRA’ MODEL OVER SAFETY CONCERNS</a></strong></p>
<p class="wp-block-paragraph">“I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we’ve had at OpenAI in recent weeks. Committing to having independent evaluators with employee-like access is a great idea, and we will do the same. We’ll have more to share soon,” Altman wrote on X in mid-September.</p>
<p class="wp-block-paragraph">The Washington Examiner has reached out to OpenAI for comment on LASST’s lawsuit.</p>