Inside the phony, incestuous web of woke AI 'watchdogs' that Anthropic claims will save us from an AI apocalypse

Key Points:

• Anthropic, an AI company, is criticized for creating a network of interconnected nonprofits and watchdog groups, including Redwood Research and METR, which are seen as lacking independence due to close personal and financial ties to Anthropic and the Effective Altruism movement.
• Redwood Research, heavily funded by Effective Altruism-linked Coefficient Giving and closely connected to Anthropic’s leadership, co-authored a controversial report on AI risks but is viewed by experts as too cozy with Anthropic to serve as an impartial AI safety overseer.
• METR, backed by Effective Altruism leaders and financially supported by Coefficient Giving, has partnered with Anthropic for AI safety evaluations, raising concerns about conflicts of interest and the credibility of self-regulation within the AI industry.
• Political figures, including Rep. Josh Gottheimer and Rep. Steve Scalise, have expressed skepticism about AI companies policing themselves, warning that such frameworks may prioritize industry interests over public safety and national security.
• Anthropic’s CEO Dario Amodei’s push for third-party AI oversight has drawn mixed reactions, with some tech leaders supporting the idea of independent evaluators, while critics, including Pentagon officials and former President Trump’s circle, question the motives and effectiveness of the current oversight ecosystem.