Understood that containment failed. But I don't think there's value in characterizing it as throwing all caution to the wind. Let's discuss how the containment failed and how to mitigate it.
The root cause of the containment failure, in the deepest sense, was that their next-generation model was better at offensive security than their humans and current-generation models were at defensive security. That problem's only going to get worse if they keep training more and more capable models.
In the API, all GPT‑5 models can accept a maximum of 272,000 input tokens and emit a maximum of 128,000 reasoning & output tokens, for a total context length of 400,000 tokens.
So it's only 270k for input and 400k in total considering reasoning & output tokens.
It's also not available in th EU, likely because of the GDPR. And IIRC the UK implemented something similar to the GDPR after the Brexit. Maybe that's why.
It's part of our process. Here's the internal timeline
T+0 Automatic comms thread created
T+1 XXX Is this a P0, do we need a status page?
@YYY
T+1 YYY Eyes on
T+4 ZZZ Yes
let's get super-generic status page up
@XXX / @YYY - you have one handy?
I see it now thx
Side note: the comments in this are all almost as valuable as the content itself. Remember when comments were a way people used to enhance content rather than attack or degrade it?
