Commentary: Anthropic's "pacing" push is the first step of regulatory capture

MilagrosMiceli · x · 2026-09-13

A Catalan-language thread sharply questions Anthropic's recent industry-wide safety coordination initiative, arguing it amounts to the first step of "regulatory capture." The author claims Anthropic positions itself as the sector's responsible leader before regulators and investors while requiring peers to adopt continuous evaluation measures that Anthropic itself defines, with no genuine external or independent oversight. The thread links this to OpenAI's 2019 GPT-2 "too dangerous to release" warning—followed by its release months later—as the original playbook, arguing moral-panic narratives serve as a strategic asset: they generate headlines, signal model superiority, and build regulatory barriers against competitors. It also notes Anthropic has confidentially filed for an IPO, framing the safety messaging against that financial context.

Original post →

More from AGI Musings

AGI Musings channel →