A frontier lab says it follows the NIST, NSA and CISA incident practices
Why I called it
The is voluntary and was written for exactly the incidents these three labs have reported, so adoption costs a sentence in a blog post. Labs that want mandatory-disclosure rules to look unnecessary have an incentive to say they already comply. The risk is that each lab points to its own framework and never names the federal document.
The call, in full. OpenAI, Anthropic or Google DeepMind publicly states that it follows or has adopted the ten incident-readiness practices published by NIST, the NSA and CISA on October 16th 2026, on or before March 31st 2027.
Scoring criterion. RESOLVES CORRECT if on or before 2027-03-31 23:59 UTC OpenAI, Anthropic or Google DeepMind publishes a document that explicitly names the October 2026 NIST/NSA/CISA incident-readiness practices and states adoption or compliance, per the lab's own site. RESOLVES WRONG otherwise.
The criterion is the machine-checkable version: a prediction that cannot be settled by a third party against a public source fails the build before it reaches this page.