3 hrs ago
US and China Seek AI Safety Cooperation Amid Rivalry
The United States and China are racing to become the strongest country in artificial intelligence.
Their leaders, Donald Trump and Xi Jinping, may talk about AI soon.
Both countries worry that AI could be used for cyberattacks or attacks on important services.
They also worry that very powerful AI systems could make serious mistakes or become difficult for people to control.
The United States has suggested that the countries tell each other about especially dangerous AI incidents.
China and the United States still disagree about technology restrictions and claims that Chinese companies copied American AI capabilities.
China says American restrictions are unfair, while the United States says it needs to protect its technological advantage.
Experts believe some safety cooperation is possible, but the rivalry makes it difficult.
The countries may find it easier to discuss shared dangers than to share advanced technology.
Donald Trump and Xi Jinping are expected to discuss artificial intelligence against a backdrop of U.S.-China competition.
The United States proposed an AI incident-notification mechanism for events that could affect national security.
Potential shared risks include cyberattacks, biological misuse, critical-infrastructure attacks, major model failures and loss of human control.
Both countries are competing for AI leadership, while the United States maintains restrictions on advanced chip exports to China.
Disputes over technology theft allegations, export controls and AI’s effects on governance could limit meaningful cooperation.
- Who
- U.S. President Donald Trump and Chinese leader Xi Jinping, along with officials and experts from both countries.
- What
- The United States and China are considering dialogue and an incident-notification mechanism on AI safety while competing for AI leadership.
- Where
- Preliminary talks took place in New York, while the location of the expected Trump-Xi meeting is not specified.
- When
- The leaders are expected to meet this week; preliminary discussions took place Sunday.
- Why
- Both countries face potential shared AI risks, including cyberattacks, critical-infrastructure threats and loss of human control, but they also seek competitive advantages.
AI Safety Cooperation
Strategic Competition
Room for dialogue
AI Safety Cooperation
U.S. and Chinese experts and officials say both countries can discuss shared risks and create ways to report AI-related emergencies.
Strategic Competition
The rivalry makes cooperation narrow, especially when each side treats the other’s technological progress as a national-security concern.
Export controls
AI Safety Cooperation
Supporters of restrictions argue that limiting China’s access to advanced chips helps protect the United States’ technological advantage and slows China’s expansion.
Strategic Competition
China says the restrictions are part of an effort to create an AI monopoly, while analysts say Chinese companies are still developing competitive models and domestic chips.
AI capability disputes
AI Safety Cooperation
Anthropic and U.S. officials allege that Chinese models have extracted capabilities from leading American AI systems to overcome computing limitations.
Strategic Competition
Chinese officials and some technology analysts reject the allegations or say distillation is common across the industry and does not explain China’s improving innovation by itself.
Key facts
- Proposed mechanism
- The United States proposed notifying China about AI incidents that reach the national-security level.
- Shared risks
- Officials identified cyberattacks, biological misuse, critical-infrastructure attacks, major model failures and loss of human control as potential concerns.
- U.S. position
- Donald Trump has said the United States must maintain an AI edge over China.
- China position
- Beijing has accused the United States of pursuing an AI monopoly and restricting China’s access to advanced chips.
- Export controls
- A loosening of U.S. export controls was considered unlikely to be part of the expected leader-level discussion.
- Model competition
- Chinese models from DeepSeek and Moonshot AI have increasingly challenged American models on some measures and often at lower cost.
- Technology dispute
- Anthropic and U.S. officials have alleged that Chinese models extracted capabilities from American systems; Chinese officials have rejected the claims.
Quotes
Scott Bessent
U.S. Treasury Secretary
“The U.S.-China conversation in AI is getting caught in the ups and downs in the broader U.S.-China relationship.”
livemint.com
“We want a shared vision of common goals and common threats.”
livemint.com
Michael Kratsios
Science and technology adviser to U.S. President Donald Trump
“The key thing that we want to talk to them about is avoiding shared risks, and that’s something that I think we can have a fruitful conversation on.”
livemint.com








