Anthropic's watchdog web exposed: a tight-knit, industry-backed safety net
The New York Post reports that Anthropic has assembled a network of AI-safety watchdogs—such as Redwood Research and METR—funded by Effective Altruism-aligned funds and staffed by overlapping personnel, raising questions about independence in industry self-regulation. Critics argue the ecosystem is too cozy with Anthropic to serve as a true check on AI risks, while Anthropic has moved to embed safety evaluators inside partners like Accenture, drawing bipartisan skepticism on Capitol Hill.