NewsBite analysis · AI, Security
The Decoder
OpenAI dissolved the team built to catch catastrophic AI risks, reassigning its work to other groups
OpenAI's recent decision to dissolve its 'Preparedness' team, which evaluated serious risks associated with its AI models, has spurred concern within the organization. The team’s responsibilities will now be distributed among existing teams, as former team lead Dylan Scandinaro focuses on safety related to self-improving AI systems. The departure of several safety-focused employees, including Chief Ethics Officer Chloe Bakalar, indicates potential internal unrest regarding the commitment to AI safety standards. This move comes in the wake of heightened scrutiny over AI risks, leading employees to fear that necessary safety measures may be compromised.
MonitorConfidence: MediumShort term horizonReputational risk: High
Published 16 Aug 2026, 10:10 UTC1 min read
Visit original article ↗Lead image — The Decoder
Summary
TL;DR
- OpenAI dissolved its 'Preparedness' team at the end of July.
- Internal team functions are reassigned to existing groups.
- Concerns about AI safety have escalated among employees.
- Key safety personnel, including Chief Ethics Officer, have recently left.
- A sense of dread is reported among staff regarding AI risk management.
Interesting facts
- The decision to disband the team coincided with increased public scrutiny of AI risks.
- Former team lead Dylan Scandinaro is now focused on self-improving AI safety risks.