Washington is preparing to receive Chinese President Xi Jinping for a visit in which artificial intelligence is expected to feature as one of the main issues between the United States and China, amid mounting economic and technological competition between the two countries. The visit coincides with the participation of leaders of several major US technology companies in an official dinner hosted by the White House in honour of the Chinese president.
Technology company leaders at White House dinner
Reuters quoted US officials as saying that the guest list for the dinner includes the chief executives of Apple, Nvidia, OpenAI, Amazon, Tesla, Google and Dell. The presence of these executives puts the technology sector at the heart of a visit in which political and trade negotiations overlap with the future of advanced technologies and the rules governing them.
The meeting comes at a time when US-China competition in artificial intelligence is no longer limited to developing chatbots or language models. The race extends to producing advanced models and chips, building data centres, attracting investment and talent, as well as drafting technology governance rules and defining controls for its use.
Gap between American and Chinese models narrows
Data from Stanford University’s AI Index 2026 show that the gap in the performance of American and Chinese models has narrowed significantly. Since the beginning of 2025, models from the two countries have alternated at the top of performance tests, while the best American model’s lead was no more than 2.7% on one of the indicators tracked by the report by last March.
Despite the convergence in performance levels, the United States still leads in the number of advanced models produced by its companies. According to the index, US companies produced 59 notable models during 2025, compared with 35 models in China. China, meanwhile, leads in the volume of research, scientific citations and patents related to the field.
These figures reflect the expansion of the race beyond commercial competition between companies. Maintaining technological superiority remains a priority in Washington’s calculations, while Beijing is working to strengthen its domestic capabilities and narrow the gap with the United States, both in the quality of models and in the research and development ecosystem surrounding them.
Data centres and chips at the forefront of competition
The ability to train and operate advanced artificial intelligence models depends on vast computing resources, as well as advanced chips, data centres and energy supplies. Digital infrastructure and semiconductor supply chains have therefore become core elements of competition between the two countries, rather than merely supporting factors in the development of technology products.
The Stanford index indicates that the United States has the world’s largest number of data centres, recording 5427 centres in the country, more than ten times the number held by any other country.
At the same time, the report shows that most of the chips used in US data centres are manufactured by a single Taiwanese company, highlighting the sensitivity of the supply chains on which the artificial intelligence industry depends. Given this reliance on chips and infrastructure, companies such as Nvidia, OpenAI, Google and Microsoft have become parties to discussions that extend beyond the development of products aimed at consumers.
Their role is increasingly linked to economic, security and international competition issues, as well as to the United States’ ability to retain its position in the global artificial intelligence market. External competition is accompanied by a dispute within the US technology sector over the appropriate pace for developing model capabilities.
Conflicting US calls over development speed and safety
Dario Amodei, chief executive of Anthropic, called on artificial intelligence companies to slow the development of advanced models. He proposed strengthening independent safety assessments and co-ordinating the efforts of leading companies to establish common standards, alongside expanding international co-operation to reduce risks.
Amodei said his call was not aimed at halting the development of artificial intelligence, but at giving safety and oversight measures enough time to keep pace with model capabilities. The call comes amid reports of artificial intelligence models being used in harmful activities, and as the number of documented incidents linked to the technology rose to 362 in 2025, compared with 233 incidents in 2024, according to the Stanford index.
But proposals to slow development run up against considerations linked to competition with China. US President Donald Trump rejected concerns about slowing the race, while Nvidia chief executive Jensen Huang said artificial intelligence development should proceed «as fast as we can». The position reflects the priority of maintaining the momentum of US development in the face of Chinese progress.
These technological disagreements are expected to intersect with political and economic negotiations during the meeting between Trump and Xi. Expected issues during the Chinese president’s visit include trade, rare earth elements and the future of artificial intelligence, while US and Chinese officials are also holding talks on artificial intelligence safety.
The White House dinner therefore goes beyond its protocol function, bringing the Chinese president together with executives whose companies’ products, investments and infrastructure are linked to the United States’ position in the technology race. The narrowing gap between American and Chinese models also places those companies’ leaders at the centre of a discussion bringing together politics, the economy, national security and the management of risks from rapid development.
The contrast between Amodei’s call to slow the development of capabilities and Huang’s demand to proceed at maximum speed shows that the debate is no longer limited to which side will produce the most powerful model. It also concerns the pace of development, the level of oversight, the safety standards that should be applied, and the ability of governments and companies to establish rules that keep pace with technological progress while competition between Washington and Beijing continues.
