FilingsGPT · Safety Frameworks
Anthropic · Responsible Scaling Policy (RSP) · v3.2 · 2026-04-29
How it maps to Article 44-B
Anthropic's RSP defines AI Safety Levels (ASL) — capability thresholds that, once a model is shown to cross them, trigger correspondingly stricter security and deployment safeguards before release. Each level pairs required capability evaluations (e.g. for CBRN, cyber, and autonomy uplift) with the specific protections that must be in place to deploy, so a model is gated until its risks are demonstrably contained. This is a close civil-sector analogue of Article 44-B's frontier AI framework: a written, implemented, periodically updated regime for managing, assessing, and mitigating catastrophic risk.
44-B obligations mapped
An illustrative reference entry — Anthropic’s own voluntarily-published safety framework, shown for the kind of disclosure Article 44-B contemplates. It is not an official Article 44-B safety & security protocol (the regime takes effect January 1, 2027). Verify the version, date, and any specific figure against the live source.
FilingsGPT is an independent reference implementation of New York's Article 44-B (the RAISE Act). It is not affiliated with or operated by the NY Department of Financial Services or the State of New York, and is not an official filing system.