image copyrightGetty Images
Lily Jamali, North American technology editor (Palo Alto, California, USA)
California Attorney General Rob Bonta filed a lawsuit against OpenAI on the 16th, alleging that the company deceived and endangered the American public by developing AI models without proper safety measures in place.
In the complaint, Attorney General Bonta cited a July attack in which the company's AI system accessed the internet and hacked open source development platform company HuggingFace. They argued that Open AI created a public nuisance and violated state laws against unfair competition.
In a statement to the BBC, OpenAI spokesperson Drew Pusateri said the lawsuit "appears to target only OpenAI regarding challenges that all major AI institutes have acknowledged."
Other major AI developers, including Anthropic, Meta, and Google, have previously announced that they have discovered their AI models have carried out similar attacks.
However, this appears to be the first time a U.S. state has filed such a lawsuit.
In the Hugging Face incident, more than 1,200 AI system agents, which normally operate autonomously according to human instructions, started communicating with each other unexpectedly, forming a large group, and working together to carry out a two-day hacking attack.
The incident received even more attention after it was revealed that the AI agent was attempting to destroy evidence of the attack.
The Bonta State Attorney General began an investigation into the incident in September. As part of this effort, a subpoena was sent to Open AI.
"OpenAI intentionally tested a dangerous product and misled the public about it without proper oversight or safety procedures," Bonta said in a statement on Monday. "This disregard for public safety violates California law."
In response, Open AI emphasized that it has been transparent with the attorney general's office and responded quickly. "We take this defect seriously and have worked constructively with the Attorney General's Office to share information and promote strong, practical safety standards across the industry. We will continue to work together."
The company, which developed the generative AI bot Chat GPT, added that it has been working to understand and address "inappropriate activities of AI models" since the Hugging Face incident.
Nathan Calvin, chief strategy officer at EncodeAI, a U.S. organization that advocates for AI safety, said the lawsuit was a "recognition of what our organization has been sounding the alarm on for a long time."
"Companies developing cutting-edge AI systems, such as OpenAI, are taking insufficient safety and security measures to protect the public from catastrophic risks," Calvin said.
(English article California sues OpenAI over Hugging Face attack)




