Eric Schmidt argues mutual pause in AI race is unfeasible

Summary

Eric Schmidt has asserted that initiating a mutual pause in the AI race between the United States and China is unfeasible due to several critical factors. He explained that while US incentives for AI development will persist, the possibility of a local lab or China circumventing these agreements remains high, as verifying compliance is extremely challenging. According to contextual facts, inspectors would need to access secret military laboratories and thoroughly review substantial amounts of code generated autonomously by computers, a process that is currently impractical. This ongoing competition between the US and China not only affects technological leadership but also plays a significant role in shaping military capabilities and daily life.

Analysis

Eric Schmidt: Eric Schmidt is the former chief executive of Google and a prominent figure in technology and AI policy discussions. In a September 2026 interview with The Economist, he explained why a mutual pause in AI development is not feasible, citing US competitive incentives, risks of cheating by labs or China, and insurmountable verification challenges with computer-generated code in secret facilities. He advocates instead for advancing AI capabilities while developing safeguards and pursuing targeted US-China treaties starting with no-surprises agreements on AI agents. Diplomacy: Recent discussions emphasize US-China treaties focused on preventing accidental escalations from AI agents over broad development pauses. Geopolitics: The United States and China are locked in a profound AI competition that influences daily life, military capabilities, and technological leadership. Verification: AI pause agreements face fundamental barriers because inspectors cannot feasibly access military labs or review codebases generated by computers.

Categories

aiai_agentspoliticstech

Related sources

View Original Tweet