Esper, known for his hawkish views, called for more funding from Congress as well as support from corporations and universities, citing American production efforts during World War II and the space race against the former Soviet Union as examples for AI development to follow.

“Mastering AI will require a similar vision, ambition, and commitment,” he said. “We need the full force of American intellect and ingenuity working in harmony across the public and private sectors. We need your leadership and vision to ensure we maintain a strategic edge.”

Esper pointed out that China is also exporting AI-powered military equipment, including military drones, to the Middle East and is using the technology to develop surveillance techniques that “strengthen its grip over its people”.

In contrast, the US wants a version of AI that “protects our fundamental beliefs of liberty and human rights”, he said.

Esper’s speech at the National Security Commission came on the same day he had a teleconference with Chinese Defence Minister Wei Fenghe, during which the two exchanged views on “building a constructive, results-oriented relationship to prevent and manage crises, reduce the risk of misunderstanding or miscalculation and enhance overall cooperation”, according to a statement by the Pentagon.

Chinese and US defence chiefs push for stable military ties and to improve cooperation

China has declared its goal to become the global leader in AI by 2030 as part of its national strategic campaign, pouring investment and research into the field.

Chinese state investment in military AI was not disclosed, but a report presented before a US congressional committee in June estimated the amount to be hundreds of millions of dollars, with a further rise in spending expected next year.

The Pentagon’s exact spending on AI is confidential, but the latest US budget proposals – currently delayed by Congressional wrangling – would see the department share in US$1 billion for general research in the field.

According to the report, China’s advantage comes in its military-civilian fusion, which it described as a “national AI team”, spearheaded by government organisations and institutions, and with participation from major businesses including corporations such as Baidu, Tencent and Alibaba, which owns the South China Morning Post.

But the US still leads in key AI development factors, with more start-ups and high quality research, according to the Centre for Data Innovation, a non-profit think tank that focuses on AI public policy.

A recent report by the National Security Commission on Artificial Intelligence – a body created last year by the US Congress – outlined its suggestions to counter Chinese AI competition, including “heightened scrutiny during the visa process for Chinese researchers with certain risk indicators, such as ties to the Chinese military”.

China army exercise puts drone defence capability to the test ‘after years spent studying US tactics’

But Jeffrey Ding, an AI researcher at the University of Oxford, noted in a research paper last year that the exact use of AI by the People’s Liberation Army is still to be determined.

“Many of these projections are largely speculative. There is not a coherent consensus of ideas on AI in warfare within the PLA,” he wrote.

“Overly broad restrictions on Chinese STEM [science, technology, engineering and medicine] students doing research in the US would hurt US AI capabilities more than they would hurt China’s AI capabilities.

“There are other places for Chinese STEM students to go, including Canada, Europe and institutions in China.”

Despite Esper’s call for more cooperation between the military and the private sector, many in the US technology sector remain wary.

Last February thousands of Google employees signed a letter to protest the firm’s involvement in a Pentagon project that used its image recognition technology, saying: “We believe that Google should not be in the business of war.”