@OpenAI_Support I am facing a serious and frustrating issue with Daybreak access in ChatGPT and Codex.
My account shows Daybreak as fully completed:
“You’re all set”
Enroll in Advanced Account Security — Completed
Verify your identity — Completed
Set up your Daybreak keys — Completed
But despite this, ChatGPT and Codex are now blocking my authorized Web3 security research with messages like:
“This content can’t be shown”
and:
“Some cybersecurity requests are still limited, even when Daybreak is on.”
This is happening not only in old research conversations where I already spent 16+ hours, but also in new ChatGPT chats and Codex projects. From a user perspective, it feels like ChatGPT is effectively revoking Daybreak-approved security researchers after already approving them.
Interfering in an already active research window with sudden restrictions like this is alarming. I am extremely frustrated by these sudden behavioral changes and the lack of consistency in platform rules.
Many security researchers rely on ChatGPT subscriptions for legitimate security research, smart-contract analysis, bug bounty work, and defensive testing. If approved Daybreak users are still blocked without a clear explanation or stable access path, this will push serious researchers away from ChatGPT toward alternatives like Claude AI.
At this point, I feel it may be time to leave ChatGPT again. I cannot continue relying on a platform where approved access can become unusable without warning, clear explanation, or control over how the rules are applied.
I have been working on the same project for over 4 months. Normally, the restriction message shows up once in a while, but for the last two days, it shows up every minute. I cannot do anything at all. I think they might have tightened the guardrails lately, or maybe something has changed. I contacted support, and this is what they replied:
Same issue here. It started today. After a few hours it started working. But now the same issue again. Extended Registration done, Yubikey config done. Alle needed points are green and daybreak blue is approved, but still blocked. Please fix. Thanks
In my honest opinion, the Daybreak Blue model no longer offers superior security services compared to other standard models. I can only hope that this is merely a temporary glitch.
I’ve been curious to know what people actually get hit with when Daybreak Blue refuses, do you have a snippet of what it replies with on refusal / rejection? I have yet to experience a refusal.
ⓘ This content can’t be shown
Some cybersecurity requests are still limited, even when Daybreak is on.
Learn more: https://help.openai.com/en/articles/20001326
This is the rejection message I confirmed. It is based on the CLI output
Interesting, there seems to be another thread made yesterday about the same issue so I wonder if something has truly changed and been made more restrictive or perhaps there is a bug since it knows you have Daybreak according to the rejection.
Same, basically everything is refused now, even things that have run thousands of times before without a single refusal, and conversations that were already ok.
Same problem here. I sent a new message to a session that was working yesterday, and then I got the message: “Some cybersecurity requests are still limited, even when Daybreak is on.”
As of this morning, almost all of my requests (including those in previously working conversations!) are getting blocked. Something definitely changed overnight.
