Big Tech's AI Slowdown: Pact or Cartel? — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI, Anthropic, Google DeepMind, and SpaceX leaders — Sam Altman, Dario Amodei, Demis Hassabis, and Elon Musk — agreed over the weekend to slow AI development via a three-step proposal covering embedded third-party auditors, domestic lab regulation, and a global slowdown pact.
- Anthropic researcher Jacob Coxon resigned and posted a public letter warning AI could 'kill us all by the end of the decade,' saying neither OpenAI nor Anthropic is 'acting responsibly' but 'racing straight to self-improving superintelligence'; his letter has been viewed over 170 million times on X.
- Daniel Kokotajlo, ex-OpenAI and now at the AI Futures Project, says the slowdown push isn't new — over 1,000 AI lab employees signed a July public letter after the OpenAI-Hugging Face incident calling for a development slowdown, and CEOs are now 'bowing to that pressure and also claiming credit for it, [although] not rightfully.'
- Critics including Sacha Haworth of the Tech Oversight Project call any voluntary framework 'regulatory capture,' saying 'we should not be letting the foxes run the henhouse,' while NYU's Nick Reese compared the gambit to social media platforms' decade-old self-regulation playbook.
- President Trump posted that AI only needs 'a STRONG AND SMART (High IQ!) PRESIDENT' as guardrails, then called Nvidia CEO Jensen Huang on stage at a conference to declare AI concerns a 'hoax,' saying 'the robots will not be taking over.'
- China's Foreign Ministry spokesperson Guo Jiakun rejected slowdown calls as 'fearmongering,' but Redwood Research's Buck Shlegeris and the Midas Project's Tyler Johnston argued global AI coordination could mirror US-Russia nuclear de-proliferation talks.
- Kokotajlo's AI Futures Project proposed giving auditors access to labs' compute budgets and pledging to decrease research compute — a concrete mechanism Amodei's essay partially echoed by calling for external auditors like METR, Apollo, and Redwood Research to embed inside labs.
Why it matters: With no binding AI regulation likely under Trump — who dismissed AI concerns as a 'hoax' from a speakerphone — any slowdown depends on voluntary commitments by the labs racing to build superintelligence. If the verbal pact converts into enforceable compute-budget limits, it could narrow the gap between frontier labs and rivals, but critics warn it risks becoming a self-serving cartel that locks in incumbents while kneecapping open source.
Ask SkimNews
