US Pitches AI Hotline to China Ahead of Xi-Trump Summit

Beijing, China: Xi Jinping begins a state visit to the United States this week carrying one of the more delicate possibilities of the Trump era on his agenda: a proposed hotline between Washington and Beijing that would let the world's two largest economies warn each other when an artificial intelligence incident becomes serious enough to threaten national security.

Chinese state media have, for now, said almost nothing about it. The official Xinhua news agency acknowledged that Vice Premier He Lifeng and US Treasury Secretary Scott Bessent discussed artificial intelligence during talks in New York on Sunday, describing the exchanges as "frank, in-depth and constructive". But Xinhua did not say whether Beijing had accepted the notification mechanism that Washington says it proposed, and He left the meeting without speaking to reporters, as did China's chief trade negotiator Li Chenggang.

The proposal, reported by Al Jazeera, is effectively an "AI hotline". It was put forward ahead of the Xi-Trump summit in Washington from Wednesday to Friday, where the two leaders are expected to consider establishing a broader US-China dialogue on artificial intelligence and where Trump has said he expects to discuss "almost everything" with the Chinese leader. Trump is expected to welcome Xi at Joint Base Andrews in Maryland.

The United States has framed the mechanism as a matter of transparency between rivals. Bessent said Washington had proposed creating a US-China AI dialogue with particular focus on national security and a notification system covering "common goals and common threats". He told reporters after Sunday's discussions that "moving from opaque to more transparency between the number one and the number two AI powers in the world is very important". Trump's science and technology adviser Michael Kratsios told Fox News that the "key thing" Washington wants to discuss is "avoiding shared risks".

Exactly what would trigger a notification has not been publicly revealed. Scott Singer, co-director of the China AI Initiative at the Carnegie Endowment for International Peace, told Al Jazeera the channel was an obvious step. "The US and China absolutely need to be able to share information with each other as threats evolve. This seems like an obvious win for both sides," he said. "The relationship needs guardrails to keep citizens of both countries safe, and this is a productive first step."

Such a channel would not be unprecedented. In 1997, Bill Clinton and Jiang Zemin agreed to create a "direct secure telephone link" between the two countries. Washington and Beijing have also maintained a military hotline since 2008, intended to give senior defence officials 48 hours to respond during potentially dangerous incidents. The limits of such channels were exposed in 2023, when the US shot down a suspected Chinese surveillance balloon: then-Defense Secretary Lloyd Austin sought a call with his Chinese counterpart and Beijing declined the request.

The timing of the new proposal reflects an intensifying contest for AI supremacy and a widening public alarm about the technology's risks. Anthropic chief executive Dario Amodei warned in an essay this month that AI capabilities are advancing so quickly that within "6-12 months" a swarm could be "capable of taking over the entire internet with a persistent botnet (potentially causing hundreds of billions of dollars in damage)", with the damage growing without guardrails. His assertions were endorsed by rivals OpenAI's Sam Altman and Elon Musk, Al Jazeera reported.

That alarm has spread far beyond Silicon Valley. An Asia Times roundup this week described what it called a "great AI freakout", reporting that a bipartisan majority of Americans now says it is worried about existential risk from AI, and that a bipartisan majority thinks AI is moving too fast, citing polling collated by Nate Silver from Politico and YouGov. The same piece described a "civil war" within the American progressive movement between a camp that says AI is dangerous, backed by Bernie Sanders, and one that calls it "fake", backed by his former communications director David Sirota. Barack Obama is telling Democrats to make AI safety one of their "tentpole" issues, the roundup said.

Trump has so far resisted calls for a slowdown. Al Jazeera noted that he has argued the United States must keep its advantage over China and has said "whoever wins AI, wins". The Asia Times roundup identified the Trump administration among five groups opposing slower AI development, alongside investors worried about their bottom line, libertarians and techno-optimists, China hawks who fear ceding the lead, and progressives who had long dismissed AI as a bubble.

The economics of the race remain lopsided. According to Stanford University's 2026 AI Index Report, cited by Al Jazeera, US companies invested $285.9bn in AI in 2025, compared with $12.4bn in China, and the US also leads in the most advanced semiconductors needed to train cutting-edge models. Yet China has shown resilience under US export restrictions: the January release of the Chinese startup DeepSeek's model, built at a fraction of rivals' cost, sent US tech share prices tumbling. Beijing also holds advantages elsewhere in the supply chain, notably in rare earth minerals used in advanced technology and in generating more than twice as much electricity as the US, which Al Jazeera said gives China access to the cheap power AI data centres require.

How quickly that race translates into economic growth is itself now the subject of a public bet described by Asia Times. Optimists in the AI industry are forecasting annual growth of 15 per cent, a rate the roundup noted China has hit only once, in 1984, since its rapid development began, and far above the 5.3 per cent that AI experts gave as their most optimistic forecast. Economists including Alex Imas and Ben Moll are more sceptical, listing constraints such as the slow diffusion of the technology, a productivity "J-curve" in which output dips after a big innovation, the difficulty of automating the physical world with robots, political barriers, bottlenecks in inputs such as chips and Baumol's cost disease. The same piece noted that AI has so far refused to kill the software-as-a-service industry, with Stripe economist Ernie Tedeschi finding that publicly traded SaaS companies' revenue has grown faster since coding agents appeared.

The Asia Times roundup also highlighted how the domestic politics of the AI debate sits within broader questions of American inequality, citing research by economists Owen Zidar and Eric Zwick showing that about 1.7 million Americans have each built a net worth of at least $10m by owning a private business, and that for every chief executive of a public company there are more than 1,000 private-business owners worth more than $25m. Their work, summarised in The Atlantic, argues that these "pass-through" business owners have shaped the tax code far more than tech billionaires, to the point that a married couple can now pass on $30m to heirs tax-free.

The summit in which the hotline proposal sits is freighted with older disputes. Trade remains central, with no lasting settlement on the tariffs that have dominated Trump's second-term relationship with China. Last month the White House accused dozens of countries of helping China illegally bypass US tariffs. Last week the US Congress passed a bill giving Trump sweeping powers to sanction Russia's crude exports and to impose steep tariffs on buyers of Russian energy, a measure Al Jazeera said will hit Moscow's biggest customers, China and India. Beijing also ranks among Iran's key economic allies, and the near seven-month-old US-Israel war on Iran is expected to feature in the talks, as are China's dominance of rare earth supplies and Taiwan.

The Ukraine war forms another layer. An Asia Times analysis this week argued that the United States has, through the Trump administration, become an important aide to the Russian war effort, most visibly by starving Ukraine of air defence, which has made Russia's ranged campaign the area where Moscow is "most effective" and likely to remain so over the winter. On land, the analysis said, Ukraine has shown the initiative through Operation Vivaldi, an offensive conducted primarily by the 3rd Army Corps whose commander, General Andrii Biletskyi, confirmed on 15 September that the operation had begun under "maximum information silence" and that its four stages were named after the seasons of Vivaldi's Four Seasons. Deputy commander Petro Horbatenko said on 19 September that around 75 square kilometres had been liberated and that "the enemy on this section is broken, disorganized". The war's economic spillover runs directly into the Beijing-Washington relationship through energy sanctions and China's role as a buyer of Russian and Iranian oil.

Against that backdrop, Carnegie's Singer argued that a dedicated AI channel could be one of the few areas where the two rivals reach an understanding. "I think establishing a permanent channel exclusively focused on AI would genuinely be a major victory. AI has been caught up in the broader ebbs and flows of the US-China relationship, and the 2024 dialogue failed because frontier AI risks got mixed in with other issues," he told Al Jazeera. "This time looks like it might be different, and having an AI risk channel is a necessary precondition to mutually beneficial cooperation."

Singer does not expect the AI contest to slow, telling Al Jazeera that competition and cooperation can coexist. What happens next, according to the sources, turns on whether Xi and Trump can convert a proposal Beijing has so far declined to publicly endorse into a working channel, while tariffs, energy sanctions, Taiwan and the war in Ukraine crowd the same agenda.