4 hrs ago
U.S. and China plan first bilateral AI safety dialogue
The United States and China are discussing a meeting about keeping advanced artificial intelligence safer.
The meeting would focus only on AI and could happen in September.
However, U.S. statements conflict about whether a meeting is planned and whether it might happen in October.
The United States wants the two countries to watch for cyberattacks carried out by AI.
It also wants AI laboratories to share information about dangerous incidents.
American officials may raise accusations that a Chinese company copied knowledge from a U.S. AI model.
Chinese officials are concerned that the United States may not have enough rules for its most powerful AI systems.
The countries hope to communicate better if an AI system causes a crisis.
Experts say even starting this communication channel could be useful, although major agreements may be difficult.
The United States and China are reportedly planning their first bilateral talks devoted exclusively to AI since Donald Trump began his second term.
Sources told Reuters the dialogue is tentatively planned for mid-September, but a U.S. Treasury spokesperson said no meeting is planned, while another Treasury statement mentioned October.
The United States wants cooperation on monitoring AI-directed cyberattacks and has proposed that American and Chinese AI laboratories share threat information and police themselves.
Washington also plans to raise alleged Chinese distillation of proprietary U.S. AI models and concerns about future Chinese systems capable of conducting cyberattacks.
The agenda, location and participants remain unsettled, although Chinese officials reportedly view the dialogue as an important outcome of the leaders’ summit.
- Who
- Officials from the United States and China, with U.S. Treasury Secretary Scott Bessent expected to lead the U.S. side; Vice Premier He Lifeng or Ding Xuexiang could lead China’s delegation.
- What
- A proposed bilateral dialogue focused on AI safety risks, cyberattacks, model security and possible cooperation.
- Where
- The location has not been determined; the leaders’ summit is due to take place in Washington on September 24.
- When
- Sources cited mid-September; U.S. statements variously said no meeting was planned or that talks might occur in October.
- Why
- Both countries are concerned about rapidly advancing AI, autonomous cyberattacks, model misuse and the need to manage a possible cross-border crisis.
Voluntary Cooperation
Regulation and Security Concerns
How AI should be governed
Voluntary Cooperation
The United States urged G20 members to take a hands-off approach and avoid AI-specific rules, while China signed the Carolina Principles, which discourage such regulations.
Regulation and Security Concerns
Chinese officials have questioned whether the United States sufficiently regulates its most advanced AI models, while U.S. officials want safeguards against cyberattacks and model misuse.
Managing AI risks
Voluntary Cooperation
The U.S. proposal emphasizes voluntary action by AI laboratories, including sharing information and monitoring threats themselves.
Regulation and Security Concerns
China has said limits on frontier AI should apply equally to Chinese and U.S. models, while Washington is concerned about alleged Chinese distillation of American systems.
Likely outcome
Voluntary Cooperation
Experts say opening a channel to share observations and jointly monitor safety incidents could be an important first step.
Regulation and Security Concerns
Other assessments expect limited results because the talks occur amid wider technological and geopolitical rivalry.
Key facts
- Proposed timing
- Mid-September according to sources; a U.S. Treasury spokesperson said no meeting was planned, while another Treasury statement said the sides may meet in October.
- U.S. lead
- Treasury Secretary Scott Bessent is expected to lead the U.S. side.
- Possible Chinese leads
- Vice Premier He Lifeng or Politburo Standing Committee member Ding Xuexiang may lead China’s delegation.
- Main U.S. proposal
- Monitor AI-directed cyberattacks and encourage U.S. and Chinese AI laboratories to share threat information.
- Additional U.S. concerns
- Alleged Chinese distillation of proprietary U.S. AI models and potential future Chinese AI systems capable of conducting cyberattacks.
- Planning status
- The agenda, location and participants remain in flux.
- Related summit
- Donald Trump and Xi Jinping are due to meet on September 24 during a summit in Washington.
Quotes
Scott Singer
Co-director of the China AI Initiative at the Carnegie Endowment for International Peace
“The Chinese side has expressed concern around whether the U.S. has sufficient regulation around the most advanced AI models. Both sides are motivated to make sure they can manage a cross-border crisis effectively.”
theprint.in
livemint.com
“We are at a tipping point where frontier agents can cause massive damage when unmonitored. Both the U.S. and China are vulnerable. This is beyond geopolitical rivalry.”
theprint.in
livemint.com
Sources
Related news
Trump-Xi Talks Expected to Address AI Guardrails
Gates Seeks Xi Talks on Global Rules for AI Risks
Bill Gates Urges Xi Jinping to Support AI Restrictions
House Democrats Press Anthropic, OpenAI Over Rogue AI Agents
Trump-backed crypto firm backs AI venture offering restricted Chinese models
US Sanctions Turkish Bank in Escalating Pressure on Iran
Bessent Predicts Oil Collapse After Iran Conflict, Lowering Bond Yields
Xi Jinping Plans Large CEO Delegation for Washington Summit
U.S.-Iran War Stalemate Deepens Amid Sanctions and Strikes
