How China Is Responding to Calls to Rein in AI

Summary

Bloomberg's Yvonne Man speaks with Artificial Analysis co-founder and CEO Micah Hill-Smith about leading AI labs' safety commitments and the global AI race. Hill-Smith says the commitments are not a broad AI pause but a move to pace frontier development and improve safety. He expects continued progress, with Chinese labs remaining competitive and leading in open-weight models. He also warns that open-weight releases, especially those strong in offensive cybersecurity, create hard-to-control risks.

• Leading US AI labs stopped short of committing to a broad AI pause.
• The commitments focus on safety work and pacing frontier model releases.
• Micah Hill-Smith expects AI progress and model releases to continue.
• Chinese AI labs are expected to keep competing and are aggressive with open weights.
• US players intend to maintain their model-intelligence lead over China.
• Open-weight releases have a different, harder-to-reverse risk profile.
• Offensive cybersecurity capabilities in open-weight models are a key concern.