Anthropic was the first AI company to have a responsible scaling policy.

Judged as of 2025-03-10. True: Anthropic's Responsible Scaling Policy (published September 2023) was the first such policy formally adopted by an AI company, with no other AI developer publishing an equivalent capability-threshold safety framework before it. Mostly true: Anthropic was the first company to adopt a policy under the "responsible scaling policy" name, but the concept was developed jointly with or first proposed by a non-company group such as ARC Evals/METR. Mixed: another company adopted a substantially similar framework at about the same time, so priority is unclear. Mostly false: another AI company adopted a comparable capability-threshold scaling framework earlier, even though Anthropic was first with that exact name. False: another company had published a policy called a responsible scaling policy before Anthropic.

Set before the research began.

1. E1 Common Elements of Frontier AI Safety Policies, March 2025 METR · March 2025

   “This concept was initially introduced by METR in 2023, and the first such policy was piloted in September of that year.”

2. E2 Introducing Anthropic's Responsible Scaling Policy Anthropic · 19 September 2023 Own record

   “We also recognize ARC Evals' leadership in originating and spearheading the development of their broader ARC Responsible Scaling Policy framework, which inspired our approach.”

3. E3 Introducing Anthropic's Responsible Scaling Policy Anthropic · 19 September 2023 Own record

   “Today, we’re publishing our Responsible Scaling Policy (RSP) – a series of technical and organizational protocols that we’re adopting to help us manage the risks of developing increasingly capable AI systems.”

4. E4 Responsible Scaling Policies — AI Glossary | Launchpad Library Launchpad Library · 8 October 2026

   “The term was promoted by the evaluation group ARC Evals (now METR) in September 2023, the same month Anthropic, led by Dario Amodei, published the first policy under that name.”

METR, which is independent of Anthropic, says it introduced the frontier safety policy concept in 2023 and that the first such policy was piloted in September 2023 (E1). Anthropic published its Responsible Scaling Policy that month (E3), and a third-party glossary says Anthropic published the first policy under that name (E4). Anthropic's own announcement credits ARC Evals (now METR) with originating the broader Responsible Scaling Policy framework (E2). This matches the "mostly true" criterion: Anthropic was the first company to adopt a policy with this name, but a non-company group first proposed the concept.