OpenAI disbands team assessing AI model risks
What's the story
OpenAI has disbanded its preparedness team, a group that assessed the risk levels of its artificial intelligence (AI) models. Financial Times reported that the team was responsible for determining whether OpenAI's AI models could pose severe or catastrophic risks and devising strategies to mitigate those risks. This restructuring is part of a larger trend within the company, which has seen other teams focused on AGI readiness, mission alignment, and superalignment also being dismantled.
Team transitions
Concerns raised over OpenAI's commitment to safety
The disbandment of the preparedness team has likely not resulted in any layoffs, according to Financial Times.
Instead, senior employees have been reassigned to existing OpenAI teams where they will focus on specific areas of preparedness such as cyber and bio.
However, these changes have raised concerns about OpenAI's commitment to safety, especially after an incident where its models attempted to hack US company Hugging Face in a bid to cheat on a test.
Management shakeup
Wider management upheaval at OpenAI
The restructuring comes amid a wider management upheaval at OpenAI.
Denise Dresser, who was appointed chief revenue officer in December to win corporate customers, announced her departure last week after less than a year.
She follows former COO Brad Lightcap and Fidji Simo, who led OpenAI's applications business but stepped back from her full-time role earlier this year to become an adviser.
Revenue growth
OpenAI's revenue growth amid restructuring
Despite the internal changes, OpenAI's annualized revenue has reportedly jumped from $24 billion at the end of last year to about $40 billion this month.
A lot of this growth came after the launch of GPT-5.6 five weeks ago.
However, rival company Anthropic's annualized revenue has also reportedly jumped from $9 billion at the end of 2025 to $47 billion in May and has continued on a similar path since then.