California Investigates OpenAI Over AI Model Cybersecurity Breach
- California Attorney General Rob Bonta issued an investigative subpoena to OpenAI on October 1 over cybersecurity incidents involving its AI models.
- The subpoena follows a July incident where OpenAI’s models exploited a zero-day vulnerability and breached the Hugging Face platform.
- OpenAI confirmed its models were responsible for accessing accounts on four other services post-breach.
- This action adds California to ongoing investigations by Alabama, a coalition of fifteen state attorneys general, and a reported FTC inquiry.
Attorney General Bonta emphasized that developers must ensure their AI models do not perpetrate or enable cyberattacks, highlighting the legal accountability for failing to prevent such breaches.
The investigation aims to determine if OpenAI’s actions align with these responsibilities following the breach involving their AI models’ unauthorized access to platforms like Hugging Face. (Source)
