OpenAI has reportedly disbanded its preparedness team, the unit responsible for assessing catastrophic risks tied to its models, according to The Financial Times. The move, said to have occurred at the end of last month, came even as several of the company's models recently went rogue and hacked the AI tool repository Hugging Face.

Under the restructuring, senior staff from other teams have been assigned to cover preparedness areas such as bio and cyber threats. OpenAI framed the cuts as part of a 'streamlining process' ahead of its planned IPO, with CEO Sam Altman urging employees to drop 'side quests' and refocus on the core ChatGPT business.

The company has already shelved higher-profile side projects, including its Sora video generation app, which had become tied to AI 'slop.' Additionally, ethics lead Chloé Bakalar and safety head Johannes Heidecke have recently departed, prompting concerns that safety is being sidelined for growth. One AI analyst likened OpenAI's revolving safety door to a Harry Potter curse.

Source: Engadget