Policy Stakeholder·Y-CBzdxiQh·Record

Policy Stakeholder Y-CBzdxiQh

Verdictunverifiable95%

1 check · 4/29/2026

1 → unverifiable

Our claim

entire record

Policy: US Executive Order on Safe, Secure, and Trustworthy AI
Stakeholder: Anthropic
Stakeholder: Anthropic
Position: support
Importance: high
Reason: Signed voluntary pre-deployment testing agreement with AISI (Aug 2024); participated in joint US-UK model evaluation; scored 80.0% compliance across 30 indicators
Source: https://www.nist.gov/aisi
Context: One of the original seven companies to sign White House voluntary commitments in July 2023,Claude models above 10^26 FLOP threshold; existing Responsible Scaling Policy aligned with EO requirements,Quietly removed Biden-era policy commitments from website after Trump revocation (March 2025)

Source evidence

1 src · 1 check

www.nist.gov/aisi resource

unverifiable95%Haiku 4.5 · 4/27/2026

NoteThe record claims 'Stakeholder: Anthropic (unknown)' with no key fields populated. The source text is about CAISI at NIST and does not mention Anthropic at all. While the source discusses CAISI's work with 'private sector AI developers and evaluators' and mentions establishing 'voluntary agreements with private sector AI developers,' it does not specifically name Anthropic as a stakeholder or partner. The claim cannot be verified or contradicted based on this source — it is simply not addressed.

Case № Y-CBzdxiQhFiled 4/29/2026Confidence 95%