AI Safety Claims Aren’t Proof of Capture

Industry leaders and researchers warning about AI risk often do so sincerely and sometimes before they have commercial interests; labeling those warnings as 'regulatory capture' misapplies the classic capture model, which describes slow erosion over time rather than rapid alignment during a high‑salience gestation phase. Treating safety proposals from insiders as automatically self‑serving risks squelching constructive governance and misdirecting public scrutiny. — This distinction matters because it changes whether policymakers and the public treat corporate calls for restraint as genuine inputs to regulation or dismiss them as rent‑seeking, thereby affecting the shape and legitimacy of AI rules.