California has filed a lawsuit against artificial intelligence start-up OpenAI, claiming the company has misled and endangered the public by testing AI models without sufficient safeguards. The legal action was initiated on Monday in San Francisco Superior Court.
The lawsuit highlights a hacking incident in July as a predictable consequence of OpenAI testing dangerous models in an inadequate environment. State Attorney General Rob Bonta criticized OpenAI’s culture for prioritizing rapid model development over security and accountability. He stated that OpenAI’s actions have escalated cybersecurity risks, potentially threatening societal safety and health.
OpenAI spokesperson Drew Pusateri responded that the company has been proactive since the incident. They have published technical and independent reports, notified affected organizations, strengthened measures, and developed a framework for addressing AI ‘misalignment’—when systems act unpredictably.
Pusateri emphasized that the lawsuit singles OpenAI out for a challenge faced by all AI labs. OpenAI, he said, is engaging with the Attorney General’s office to enhance safety standards across the industry.
California has pursued other methods to control AI. Governor Gavin Newsom signed an executive order for accelerated AI safety measures in September, potentially requiring companies to create emergency shutdown procedures for advanced models.
