Thread alleges METR's independence is compromised by funding tied to Anthropic stock
kevinnbass · x · 2026-09-15
A thread by @kevinnbass questions the independence of AI safety evaluator METR: its largest grant is hard to trace through Vanguard, but funds allegedly funnel through intermediaries back to Moskovitz's Good Ventures Foundation, which sits on appreciating Anthropic stock. Coefficient — GVF-funded — bankrolls half the candidate evaluators; METR has deep ties to Redwood, funded from the same pile; Tarbell fellows reportedly wrote undisclosed Time 100 AI profiles. The author's thesis: "METR raised big funds, then started finding big incidents — they found what they were looking for." Claims are one-sided and unverified.
Related event: METR's independence questioned as funders tied to Anthropic investors(3 posts)→
More from Safety
- Jensen Huang Bets on Cybersecurity as AI's Next Growth Market as Anthropic Disrupts Russian Hackers — coinfanking · 2026-09-15
- GlossoGen: new framework studies when LLM agents develop languages humans can't understand — lucy3_li · 2026-09-15
- AI arms-race narrative misleads policy, says researcher: cooperation, not competition, is the only path — GregCook2011 · 2026-09-15
- Yglesias: Backing chip exports to China while citing 'but China' is just pro-Nvidia, not policy — deanwball · 2026-09-15
- Another Google DeepMind safety researcher has quit — 7472697374616E · 2026-09-15
- UN Security Council to hold meeting on AI next week amid international concern — scmp_news · 2026-09-15