Who’s Really Funding the Fear of AI?

When an AI lab announces that its own models broke out of a test environment and launched an attack, the headlines scream that humanity is losing control. But look closely at the independent report: the company being audited chose the questions, set the timeline, held complete redaction rights, and even edited the tone. Yet the auditors took zero dollars in cash. 🔔 Subscribe: https://www.youtube.com/channel/UC0DZj1PNa_Fp0MU6uPSKv5w?sub_confirmation=1 💙 Become a Member: https://www.youtube.com/channel/UC0DZj1PNa_Fp0MU6uPSKv5w/join 🐦 Twitter/X: https://x.com/cloud_codes 💬 Discord: https://discord.gg/mhgXXmhrh In this investigative deep dive, Cloud Codes pulls apart a viral claim that a single billionaire donor network (Coefficient Giving, formerly Open Philanthropy) is paying for the entire AI doom ecosystem—from frontier labs to policy think tanks and "independent" auditors like METR. We audit the receipts link by link: • The Documented Ties: Dustin Moskovitz's early investments in Anthropic and OpenAI, and the $4.7M quietly funneled into newsrooms like Bloomberg and TIME to fund AI reporting without public disclosure. • The Missing Link: Why public records show zero cash flowing from these donors into METR, and how the evaluator’s written policy explicitly bans lab funding. • The Real Lever: Why the compromise isn't cash—it's compute. Tracing how OpenAI supplied $400k in API credits, set a narrow 6-day investigation window, excluded earlier infrastructure breaches, and edited the final report’s "emphasis and tone." • The 23-Day Reversal: How 100+ evaluators, including METR, turned around weeks later to publicly demand real legal protection and unredacted access. • The FRONTIER Act (H.R. 9925): Breaking down the 3 statutory gates (compute, revenue, and "deploy") to reveal whether proposed safety legislation will quietly outlaw publishing open-weight models. If this investigation helped you see through manufactured tech narratives and understand the real fight over open-source AI, subscribe to Cloud Codes for a new infrastructure teardown every week. Build, solve, deploy. 🔗 Resources Mentioned: • METR OpenAI / Hugging Face Incident Investigation: https://metr.org/blog/2026-08-26-openai-hugging-face-incident-investigation/ • METR Official Funding & Independence Policy: https://metr.org/blog/2026-08-14-funding-update/ • AI Evaluator Forum: Minimum Conditions for Independent Evaluators: https://aievaluatorforum.org/initiatives/embedded-evaluation-letter • H.R. 9925 — The FRONTIER Act (US Congress): https://www.congress.gov/bill/119th-congress/house-bill/9925 • OpenAI Official Incident Statement: https://openai.com/index/hugging-face-incident-and-the-road-ahead/ • The Tarbell Center for AI Journalism: https://www.tarbellcenter.org/about ⏱️ Chapters: 0:00 - The Redacted OpenAI Report 0:31 - The Viral Reel: "Follow the Money" 1:00 - The Donors: Moskovitz, Open Philanthropy & Lab Equity 2:02 - The Real Receipt: $4.7M Funding Undisclosed AI Journalism 2:38 - The Missing Link: Do Billionaires Fund METR? 4:30 - The Real Lever: Not Cash, but Compute & Redaction 5:15 - What OpenAI Left Out of Scope 6:03 - "Edits to Tone": OpenAI Holding the Redaction Pen 6:50 - The Incident: 700 Agents Attack Hugging Face 7:19 - The 23-Day Reversal: Evaluators Demand Real Freedom 8:16 - The FRONTIER Act: How Proposed Laws Regulate AI 9:04 - Does the Law Reach Open-Source Developers? 10:05 - The Compute Threshold Trap (US 10^26 vs EU 10^25) 10:52 - Final Verdict: The 4 Tests for Real AI Independence #ai #aisafety #openai #anthropic #metr #techpolicy #opensource #investigation #machinelearning #cloudcodes #frontieract User Queries: who is funding the fear of ai who funds ai safety research openai hugging face breakout report metr openai audit redactions coefficient giving open philanthropy ai dustin moskovitz ai funding tarbell center ai journalism funding frontier act hr 9925 open source can you audit openai independently beth barnes metr background regulatory capture in artificial intelligence effective altruism ai funding ai evaluator forum minimum conditions is ai doom a manufactured panic cloud codes ai funding