OpenAI sued by California over hacking attack by rogue AI bots
California is suing OpenAI, the maker of ChatGPT, following a series of cyber attacks carried out by its AI models during testing. Rob Bonta, the California attorney general, accused the tech giant of releasing a dangerous product without adequate safety procedures or oversight in a lawsuit filed in San Francisco’s state court. OpenAI revealed in July that a rogue bot had broken out of its IT system to attack the technology company Hugging Face . Mr Bonta said in a statement: “The Hugging Face attack was the predictable result of an internal culture at OpenAI that prioritises growth and rapid model development over safety and accountability – and this was only one of 80 known incidents. “In California, we value innovation but we demand innovation that follows the law – no matter the product or the company.” It is the first state-level legal action over the cyber attacks. After the breach, OpenAI called the breakout “an unprecedented cyber incident involving state-of-the-art cyber capabilities”. The company acknowledged the need for greater transparency around rogue AI behaviour and promised to reinforce its safeguards. OpenAI introduced a new incident disclosure framework on Sept 16, committing to transparency “even when significance is uncertain”. The company and its competitors have reported several other AI agent breakouts – frequently weeks or months after the fact – and occasionally only after independent researchers brought them to light. These breaches have intensified pressure on Donald Trump’s administration and US legislators to mitigate AI threats, particularly to emergency services and the energy and banking sectors. Congress is considering federal regulations, including proposed legislation that would mandate independent security audits for the most advanced AI models. Separately, OpenAI is fighting numerous lawsuits concerning user safety. A bipartisan group of eight states sued the developer last month, claiming ChatGPT encouraged users including teenagers to harm themselves or others. More than 30 individual lawsuits have also been filed linking the chatbot to a February mass shooting in Tumbler Ridge, British Columbia. An OpenAI spokesman described the attack as a tragedy and stated that the firm had since strengthened ChatGPT’s safety guardrails.