Ex-Anthropic Whistleblower Suspects Chinese Spies Inside America’s AI Labs
October 6, 2026
AI Summary
5 min readA former researcher at OpenAI and Anthropic, Jacob Coxon, resigned in September 2026 with a single tweet that received 175 million views and prompted a White House meeting on AI safety. He left behind a seven-figure compensation package because he believes the companies are racing toward self-improving superintelligence without adequate safeguards. Coxon argues that 2027 will be a "crazy year" when AI systems become capable enough to automate AI research itself, triggering a runaway race that private companies are ill-equipped to manage.
The case for slowing down
Coxon's central claim is that frontier AI labs are treating a "super weapon technology" like a normal business. He draws a direct analogy to nuclear weapons: "You wouldn't let Mark Zuckerberg have nukes." The problem, as he sees it, is that free-market incentives push companies to prioritize speed over safety. When a lab must choose between allocating compute resources to making the model smarter versus doing safety research, the competitive pressure to stay ahead—especially of China—tilts the decision toward speed.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of PBD Podcast
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:02) **Opening: The Core Thesis** - The host frames the episode around Jacob’s explosive resignation tweet and the central claim that AI is a super-weapon being recklessly developed.
- 2 (02:03) **The Whistleblower's Entrance** - Jacob Coxon is introduced, and the story of his viral resignation from Anthropic is laid out.
- 3 (03:23) **Why He Quit: The 2027 Prediction** - Jacob explains the specific timeline and technical reason for his fear: the automation of AI research itself.
- 4 (05:20) **Nationalization vs. Free Market** - A debate on whether AI should be controlled by the government or private companies.
- 5 (08:48) **Addressing the "Fear Porn" Argument** - Jacob defends his motives against claims he is engaging in regulatory capture or hype.
- 6 (10:18) **The Chinese Spy Problem** - Jacob asserts with 90% certainty that China has spies inside OpenAI and Anthropic, citing a lack of proper vetting.
- 7 (13:08) **How China Catches Up: Distillation** - Jacob explains the mechanism by which China benefits from US AI development.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
Former OpenAI and Anthropic researcher Jacob Coxon explains why he quit the AI race, why he believes 2027 could be a turning point, how Anthropic differs from OpenAI, and why self-improving AI could upend jobs, national security, markets, and who ultimately controls global power in the years ahead.
⠀
⠀
👞 Get the NEW Billionaire Suede Loafer:
https://bit.ly/4ym5oRA
⠀
🍫 Get The FLB 1's NOW IN CHOCOLATE:
https://bit.ly/4gXlv0X
⠀
👕 SHOP THE ICON COLLECTION:
https://bit.ly/3TuUU3x
⠀
Ⓜ️ CONNECT ON MINNECT:
https://bit.ly/4kSVkso
⠀
Ⓜ️ PBD PODCAST CIRCLES:
https://bit.ly/4mAWQAP
⠀
👔 BET-DAVID CONSULTING:
https://bit.ly/4lzQph2
⠀
🥃 BOARDROOM CIGAR LOUNGE:
https://bit.ly/4pzLEXj
⠀
🇰 KALSHI:
http://kalshi.com/pbd
⠀
⠀
SUBSCRIBE TO:
⠀
@VALUETAINMENT
@BetDavidConsulting
@vtgenz
@ValuetainmentComedy
⠀
⠀
ABOUT US:
⠀
Patrick Bet-David is the founder and CEO of Valuetainment Media. He is the author of the #1 Wall Street Journal Bestseller “Your Next Five Moves” (Simon & Schuster) and a father of 2 boys and 2 girls. He currently resides in Ft. Lauderdale, Florida.
More from this podcast
PBD Podcast →