LIVE
All stories ›
AI IN LIFENEWS
Tools & AppsBusiness & DealsAI ModelsResearchSocietyChips & ComputeSafety & SecurityRegulation & PolicyRoboticsReviews OpenAIAnthropicGoogle & DeepMindAlibaba / QwenMetaxAIByteDance
Home › Business & Deals › BUSINESS
BUSINESS

US and China Open AI Hotline for Incidents

During Xi Jinping's state visit to Washington, both sides set up a direct line for AI incidents; the next round of talks is due by November.

US and China Open AI Hotline for Incidents
Symbolic illustration: a hand lifts the red handset of a crisis line in a situation room as a warning lamp flares beside it.

In short

The United States and China have opened a dedicated communication channel for incidents involving AI systems, agreed during Xi Jinping's state visit to Washington.

At a glance

  • New channel for AI incidents between Washington and Beijing, agreed during Xi Jinping's three-day state visit.
  • The next session in this format is to take place by November.
  • Both sides will use Trump's preferred term “superintelligence” instead of “artificial intelligence”.
  • Alongside it: tariff cuts on goods worth $30 billion in each direction.
  • China is to buy 10 million tons of US coal per year for two years.

The United States and China now have a dedicated way to reach each other when something goes wrong with an AI system. The two governments set the channel up during Chinese leader Xi Jinping's three-day state visit to President Donald Trump in Washington. It is meant to carry a recurring conversation about the technology's risks and its benefits. A follow-up session is due by November.

A standing line, not a one-off meeting

What the two sides agreed is a channel rather than a single conversation: the talks are supposed to recur on a schedule. Handelsblatt reports the agenda covers both the dangers and the upside of the technology. Which agencies staff the line, and what actually triggers a call during an incident, has not been disclosed.

Washington also got its vocabulary

The visit settled a naming question as well. Both delegations agreed to say “superintelligence” where officials would normally say “artificial intelligence”. The term is Trump's; his stated reason was that the conventional phrase makes the technology sound “fake”.

Bundled with a trade package

The AI channel landed next to the visit's commercial results. Each side is cutting tariffs on $30 billion worth of goods moving the other way — US farm products, timber, cosmetics and medical technology in one direction, small household appliances, toys, holiday decorations and car seats in the other. China also committed to buying 10 million tons of American coal a year for the next two years.

What is not confirmed here

Other coverage framed the mechanism as a “red telephone” for AI and credited UN Secretary-General António Guterres with calling for it. That attribution comes from a Spiegel summary line; neither the full Spiegel piece nor the Zeit Online report could be retrieved for this article, so it is not independently verified here. Everything above rests on the Handelsblatt account linked below. Whether a scheduled channel moves fast enough for a live AI failure is a question the agreement does not answer.

◈ AI-GENERATED REPORT · SOURCES LINKED

FAQ

What did the US and China agree on AI?

A communication channel for incidents involving AI systems, through which both governments intend to discuss the technology's risks and benefits on a recurring basis.

When is the next US-China AI talk scheduled?

By November, according to the agreement. No exact date was given, and the participating agencies were not named.

Why say superintelligence instead of artificial intelligence?

Both sides adopted Trump's preferred wording. He argued that “artificial intelligence” makes the technology sound fake.

Sources

More reports