WorkSeptember 12, 2026via The Decoder
Anthropic CEO Amodei wants AI speed limits before self-improvement outpaces human control
Why it matters
A frontier lab leader is publicly advocating for regulation and industry coordination on AI development velocity, framing safety governance as a competitive necessity rather than a constraint. This signals a shift in how labs position themselves to regulators and the public—and lands just before a potential record IPO, suggesting governance and policy risk are now IPO-material.
Key signals
- Dario Amodei proposes embedded auditors at AI companies
- Calls for shared safety standards and global agreements modeled after SALT disarmament treaties
- Warns recursive self-improvement could threaten internet within 6-12 months
- Statement precedes potential largest IPO in history
- Policy/governance positioning as part of public narrative ahead of capital raise
- Dario Amodei (Anthropic CEO) calls for controlled slowdown in AI development
- Proposes embedded auditors at AI companies
- Proposes shared safety standards across industry
- References SALT disarmament treaties as model for global AI agreements
- Timing: announcement precedes potential largest IPO in history
The hook
Anthropic's CEO just called for AI speed limits—embedded auditors, global treaties, the works. Here's why he's saying it now.
Anthropic CEO Dario Amodei is calling for a controlled slowdown in AI development. He warns that recursive self-improvement could threaten the entire internet within six to twelve months and proposes embedded auditors at AI companies, shared safety standards, and global agreements modeled after the …