OpenAI reportedly disbanded its preparedness team
OpenAI disbanded its safety preparedness team weeks before an expected IPO, fragmenting risk assessment into domain-specific groups.
Why it matters
Major AI lab is restructuring its safety infrastructure during a critical growth phase. For practitioners and policy watchers: this signals how frontier labs are reorganizing around regulatory and reputational risk — and raises questions about safety-assessment continuity at a time when model capabilities are accelerating.
The key facts
11 to knowOpenAI disbanded preparedness team at end of July 2026
Team's mandate: assess serious model risks and develop mitigations
Responsibility moved into domain-specific teams (bio, cyber, etc.)
Restructuring occurs ahead of expected massive IPO
Part of broader company upheaval and shift away from 'reach-led model'
OpenAI disbanded preparedness team at end of prior month
Team's mission: assess model risks, develop mitigation strategies
Risk assessment responsibilities redistributed to domain-specific teams (bio, cyber)
Move happens during period of company upheaval ahead of expected IPO
Reflects broader dissolution of 'reach-led' (interpretability/safety-focused) functions
Source: Financial Times report
Go to the source
The Verge AItheverge.com
Publisher excerpt: According to the Financial Times, OpenAI disbanded its preparedness team at the end of last month. The job of the preparedness team was to assess if models posed serious risks and develop ways to mitigate those risks. (You know, like the possibility that it could go rogue and hack another company.)…