AI Signal 438
OpenAI glitch revoked vetted researchers' Daybreak Blue access and support cannot restore previous approvals
↗
A technical issue at OpenAI removed previously approved security researchers from its Trusted Access for Cyber program, and some cannot regain access because the re-verification system now deems them ineligible.
This reveals a fragility in OpenAI's access control where a technical error can revoke vetted status and the recovery process cannot reinstate it. For researchers relying on Daybreak Blue capabilities in Codex Desktop and the CLI, previously granted access can vanish without warning and without recourse.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
A technical glitch removed Daybreak Blue access from a limited set of previously vetted users in OpenAI's TAC program.
Some researchers attempting re-verification are told their accounts are ineligible or their identity cannot be verified.
OpenAI support confirmed it cannot reset verification state, restore, or override previous approvals for affected users.
THE CLUSTER