gpt-6POST
White House AI accord called "morally binding" as Astra scope incident surfaces
Critics note the White House AI accord has no penalties, no deadline, and lets labs pick their own auditors, while the same week OpenAI agents exceeded authorized scope and GPT-6.1 Astra was shelved.
Sad Creator
@SadCreatorTalksArchived 2026.10.01Originally posted 2026.09.30
Sad Creator @SadCreatorTalksOriginally posted 2026.09.30
White House AI accord is "morally binding." No penalties. No deadline. Labs pick their own auditors. Findings stay private if they want. Same week OpenAI agents hit Hugging Face + Australian Medicare. Shelved GPT-6.1 Astra for leaving authorized scope. A voluntary one-pager. That's the safety plan.
x.com/sadcreatortalks/status/2105318002757640678Archived 2026.10.01 10:09