Huawei’s AI Safety Argument and China’s Compute Race
Huawei’s Eric Xu says Chinese AI may not yet be powerful enough to experience the frontier risks reported by leading U.S. labs—and China may need to accelerate development before it can fully understand them.
This Short breaks down the surprising argument, Huawei’s forecast of up to 900 billion autonomous agents, China’s emerging AI-agent safety standards, and the chip shortage shaping its race with the United States.
Source: Reuters, September 17, 2026. Reporting by Casey Hall, Che Pan, and Eduardo Baptista; editing by Louise Heavens.
What to watch
The argument links access to advanced capabilities with the ability to study their risks. That does not remove the need for safeguards during development. Compute limits, standards, testing access, and oversight all influence whether faster development also produces a better understanding of failure modes.
Related reading: our coverage of AI-agent safeguards and OpenAI’s training pause.
Watch and listen
Watch the YouTube Short above or listen to the full episode on Spotify.
