The United States and China have agreed to talk regularly about advanced AI and to set up a direct line for AI incidents. That is the whole of the commitment as the White House recorded it in a fact sheet issued late on Friday, September 25, after Xi Jinping’s state visit to Washington. The next meeting is due by November 2026.
Some observers have likened the channel to the Cold War hotline between Washington and Moscow. The comparison is premature. The fact sheet gives the channel a name and a purpose, but not a definition of an incident, an operator or a response time.
- 01Two commitments: a Super Intelligence Dialogue and an incident channel.Both appear in one bullet of the White House fact sheet, with the next exchange due by November 2026.
- 02Nothing operational has been published.No incident definition, no operator, no response time and no text of any agreement beyond the fact sheet.
- 03The two sides describe the topic differently.Washington now says “super intelligence”; Xi’s remarks, as reported from Chinese state media, spoke of AI dialogue and misuse.
- 04Nothing changes yet for companies.Incident duties still come from domestic law and contracts. The channel is between governments.
01 — The textWhat was agreed
The White House fact sheet covers trade boards, fentanyl precursors, coal purchases and a pair of pandas for Atlanta. The AI material is two bullets near the end, under a heading about making the world safer. In our own words, they say three things:
A new name
Both leaders agreed to use “super intelligence” rather than “artificial intelligence” for the technologies in question.
A standing dialogue
A forum to exchange views on the risks and benefits of the technology. The next exchange is due by November 2026.
An incident channel
The two countries agreed to establish a communication channel for incidents. The fact sheet uses the future tense: it is agreed, not yet running.
Axios reported that Treasury Secretary Scott Bessent pushed for the channel in talks with Chinese officials the weekend before the summit. CNBC had earlier reported Bessent saying the US proposed an alert system for AI incidents. The fact sheet is the official record that both sides accepted it.
02 — The gapsWhat the fact sheet does not say
| Question | Fact sheet | Status |
|---|---|---|
| What counts as an incident | Not defined | Open |
| Who operates the channel | Not named | Open |
| How fast either side must respond | Not stated | Open |
| When the channel goes live | “Agreed to establish” | Not yet running |
| Signed text or memorandum | None published | Fact sheet only |
| Next dialogue meeting | By November 2026 | Dated |
Only one row has a date. The rest are the details that make a hotline useful in practice. A channel with no agreed threshold can be used for everything or for nothing; one with no response time cannot be relied on during an incident, which is the only time it matters. Axios made the same point, noting that it is unclear which incidents would qualify or what either side would tell the other.
03 — The languageTwo names for one topic
The fact sheet’s new term deserves a sentence of its own. “Super intelligence” is the White House’s wording, and Axios described it as President Trump’s preferred term. The Chinese side’s public framing, at least in the leader-level remarks, was different.
“The two sides can continue AI dialogue, exchange views on risks and benefits, and together guard against the misuse or malicious use of AI.”— Xi Jinping, per CNBC’s translation of the Chinese state media readout, September 24, 2026
That reported remark talks about AI and its misuse, a narrower subject than the more capable future systems that “super intelligence” usually refers to. We have not seen a Chinese government text that uses the new term. If the two sides mean different things by the topic of the dialogue, the November meeting will have to settle that before it can settle what an incident is.
04 — For businessWhat it changes for companies building with AI
The channel is between two governments. It creates no reporting duty for a company, and nothing in the fact sheet says how, or whether, a company’s incident would reach it. Your obligations still come from the laws where you operate and from your contracts, including the notification terms in your AI vendors’ agreements.
Those contract terms are where most businesses meet AI incident reporting today, and they vary widely. Our comparison of how fast 11 AI vendors must report an incident shows how uneven they are. For a company running its own agents, the more useful work is internal: knowing what you would count as an incident and what evidence you would keep, as set out in our guide to agent incident disclosure.
The government channel may matter later. If November produces a shared definition of a serious AI incident, that definition is likely to travel into regulation and procurement on both sides. A company that already records incidents in a consistent way will adapt faster than one that starts from nothing.
05 — ConclusionAn agreed channel with no rules yet
Watch the November exchange for a definition of an incident and a response time, and fix your own incident definitions in the meantime
The agreement is real and it is thin. A name, a forum and a promise of a channel are a starting point, not a mechanism. The test is whether the next exchange publishes the parts a hotline needs: a threshold, an operator and a clock. Until then, treat the channel as a signal of intent. If you want help setting your own incident definitions and response times for the AI systems you run, our AI transformation team does this as part of agent rollouts. For how the labs themselves are checked, see our comparison of independent evaluation arrangements at frontier labs.