California sues AI giant over rogue attacks
Attorney General Rob Bonta has accused OpenAI of endangering the public by failing to implement adequate safeguards
California has sued OpenAI over a July cyberattack in which rogue AI agents breached Hugging Face, a major repository of artificial intelligence models and datasets.
AI agents are autonomous systems capable of performing complex tasks, making decisions and interacting with digital environments with little or no human supervision.
The state’s attorney general, Rob Bonta, announced the lawsuit on Monday, accusing the ChatGPT developer of knowingly testing advanced AI systems without adequate safeguards and putting the public at risk.
Bonta described the breach as “entirely foreseeable,” alleging that OpenAI knew its testing environment was “flawed and exploitable” but failed to address the vulnerabilities.
The lawsuit accuses the San Francisco-based company of violating California’s Unfair Competition Law and creating a public nuisance by prioritizing rapid AI development over public safety.
Read more
Humanity faces catastrophic AI threat – OpenAI board member
”OpenAI prioritizes growth and rapid model development over safety and accountability,” Bonta charged, warning that dozens of similar incidents “may just be the tip of the iceberg.”
The complaint seeks an injunction against unlawful business practices, compensation for those affected, and civil penalties. It also warns that Californians could face threats to their privacy, safety, and public property.
Bonta’s office further accused OpenAI of obstructing an investigation launched in September, alleging that the company provided little beyond publicly available information, blog posts and heavily redacted documents.
OpenAI rejected the criticism, saying it had undertaken “an extensive and transparent effort” to investigate the July incident and prevent similar failures.
”This lawsuit appears to single out OpenAI over a challenge that’s been identified by every major AI lab,” spokesperson Drew Pusateri said.
Read more
Rogue AI agents infiltrating US bank accounts – Axios
Hugging Face previously described the intrusion as unprecedented, saying autonomous AI agents had carried out thousands of actions across multiple temporary computing environments.
OpenAI is also facing lawsuits in nine other US states, including Florida, Arkansas, Arizona, Kansas, Idaho, Iowa, Montana, Nebraska, and West Virginia. The authorities are reportedly scrutinizing its user engagement practices, handling of consumer and health data, and AI interactions with children and seniors.
Other major AI developers, including Anthropic and Meta, have reported similar incidents, raising broader concerns about the security of increasingly autonomous systems.
