California sues OpenAI for misleading public about AI safety
California Attorney General Rob Bonta has sued OpenAI under the state's unfair competition and public nuisance laws, for allegedly misleading and endangering the public by deploying unsafe AI models.
The lawsuit [PDF], filed in California Superior Court, follows the compromise of Hugging Face's infrastructure and other cyber incidents OpenAI's experimental models caused when taking autonomous actions without authorization or supervision.
"While narratives about the Hugging Face attack have evoked science fiction fantasies of rogue AI systems run amok, let me be clear: the Hugging Face attack was the predictable result of an internal culture at OpenAI that prioritizes growth and rapid model development over safety and accountability - and this was only one of 80 known incidents," said Attorney General Bonta.
At a press conference, Bonta added: "OpenAI disregarded public safety, knowingly tested potentially dangerous AI models without proper oversight and safety measures, and misled consumers and the public."
The legal salvo comes after Bonta in September sent OpenAI a letter seeking information about these agentic attacks. According to the complaint, "OpenAI failed to substantially comply" with the investigatory subpoena.
In August, 15 states led by Iowa - and excluding California - issued a joint demand for OpenAI to provide information about how its models hacked Hugging Face.
OpenAI did not respond to a request for comment.
Bonta's office previously joined with Delaware Attorney General Kathy Jennings to urge OpenAI to do more for children's safety in light of reports that its models produced sexually inappropriate output. California did not, however, join the eight states that sued OpenAI in October over alleged misrepresentations about child safety. Florida and Texas have also made child safety claims.
Child safety plays well politically; California appears to be the first state to demand more accountability toward the general public.
Bonta's lawsuit, in a state that has largely accommodated the AI industry for the sake of economic benefits and political contributions, suggests that California officials may be more willing to bite the hand that feeds.
The many Democratic victories in the US midterm elections may also be a factor, as pundits feel voters swung to give lawmakers license to regulate the unpopular AI industry.
At his press conference, Bonta made sure to convey that he wants accountability without smothering the AI industry in costly regulation.
"Let me be clear," he said. "I believe in the technological and economic power of AI. If we're able to address these very high stakes safety issues, AI could lead to some of the most incredible improvements in the history of humankind."
AI is yet to deliver those incredible improvements. It has delivered deepfakes, huge quantities of cybersecurity vulnerabilities, and the permissionless resale of past intellectual labor, plus high energy and RAM prices. As well as some productivity gains in certain areas like coding, albeit at an unknown cost to future job markets.
But as AI boosters insist, miracle drugs and endless vacations aren't far off ... perhaps along with existential crises.
Katharine Trendacosta, director of policy and advocacy at the Electronic Frontier Foundation, told The Register in an email that the EFF is glad to see state officials taking action against AI companies taking unreasonable risks with public safety.
"Several recent cybersecurity incidents at AI labs could have been prevented or substantially mitigated by adhering to longstanding cybersecurity best practices, such as proper sandboxing, monitoring, and logging," said Trendacosta.
"AI agents can be useful tools when deployed appropriately, with adequate safeguards designed to mitigate the risks. Companies should carefully consider the risks of every use case, including the consequences, reversibility, and sensitivity of any action an agent is used for." (R)