US Treasury Secretary Scott Bessent has confirmed that artificial intelligence is set to feature prominently when President Donald Trump and Chinese President Xi Jinping meet in Washington this September, framing the two nations as the world's dominant AI powers with a shared obligation to establish safeguards around the technology.

What Bessent Said

Speaking to reporters on the sidelines of the G20 finance ministers' meeting in Asheville on 2 September, Bessent explained that substantive AI discussions with the Chinese delegation were limited at the summit because Beijing's designated AI interlocutor — Vice Mayor He Lifeng, Bessent's direct counterpart — was not present.

'We had limited discussions with the Chinese on AI because the designate for their AI discussions is my counterpart, Vice Mayor He Lifeng. So he was not here,' Bessent said. He indicated the planned Trump-Xi summit would serve as the primary venue for more consequential dialogue on the subject.

'I expect that when President Trump and Xi Jinping meet in Washington in September, that AI will be on the agenda,' he said.

The Case for Guardrails

Bessent argued that as the 'number one and number two superpowers in AI, by a lot', both countries bore a special responsibility to manage the technology's risks. A key concern he raised was the prospect of advanced AI systems — specifically large language models — falling into the hands of non-state actors.

'It behoves all of us to put up guardrails, to have close communication,' he said, adding that the two nations needed 'to prevent the non-state actors from achieving any of these powerful, large language models.'

Bessent did not specify what form those guardrails might take — whether the discussions would encompass export controls, computing infrastructure standards, AI safety benchmarks, or access restrictions on frontier models.

Broader G20 Conversations

Beyond the US-China dynamic, Bessent said Washington held more extensive AI discussions with other G20 member nations, who he described as keenly interested in the United States' approach given its leading position in the sector.

Those conversations covered how financial institutions are preparing for AI-related risks and opportunities, including the resilience frameworks and industry working groups that US regulators have established. 'We are happy to show them how we have structured industry working groups on this,' he said.

The G20 chair's statement described AI as a general-purpose technology capable of reshaping the global economy, noting that countries that responsibly develop and spread the technology would likely lead global growth in coming years. The document also acknowledged financial and sector-specific risks, backing regulatory frameworks aimed at preserving financial stability and strengthening cyber resilience.

Context and What Comes Next

Trump and Xi have engaged repeatedly on trade, technology, and security across Trump's two terms. The forthcoming Washington meeting would mark the first occasion where AI governance is reportedly expected to be addressed directly at the heads-of-state level between the two powers.

With AI models advancing rapidly and their integration into financial systems deepening, the outcome of the Trump-Xi AI discussions — and whether any concrete framework emerges — will be closely watched by governments and industry bodies worldwide.