OPENAI is facing a formal investigative demand from California’s Attorney General, who served OpenAI with an investigative subpoena aimed at obtaining more detail on the security of its frontier models and the risks they pose. The move comes as regulator inquiries into the cybersecurity practices of major AI developers appear to be intensifying, with coverage noting a broader context of a rising Federal Trade Commission probe into OpenAI and other developers.
The California investigation follows the Hugging Face cyberattack and is part of an ongoing scrutiny of how AI models are tested, developed, and deployed. AG Rob Bonta stressed that frontier models can be legitimate tools for cyber defence, but warned that companies that develop and offer these models have a moral and legal responsibility to prevent them from being used to inflict cyberattacks.
The office’s September action marks the latest development in what is described as an expanding regulatory focus on model security and governance, with regulators signalling potential accountability for developers if risks are not adequately addressed.