US Lawmaker Presses Chinese AI Firms Over Catastrophic AI Risks

Washington — U.S. Rep. Ro Khanna has warned that neither the United States nor China would benefit if humans lose control of advanced artificial intelligence, as he pressed Chinese AI companies and the U.S. intelligence community for safeguards against potentially catastrophic biological, nuclear and economic risks.

Khanna, the top Democrat on the House Select Committee on the Strategic Competition Between the United States and the Chinese Communist Party, sent letters to DeepSeek, Alibaba and Moonshot AI seeking information about their plans for artificial general intelligence and superintelligence.

“Neither the US nor China wins if humans lose control of this technology or misaligned AI triggers a global economic, biological, or nuclear crisis,” Khanna wrote.

The letters were addressed to DeepSeek CEO Liang Wenfeng, Alibaba CEO Eddie Wu and Moonshot AI CEO Yang Zhilin.

Khanna also wrote separately to Director of National Intelligence Jay Clayton seeking two formal intelligence assessments related to the risks posed by increasingly powerful AI systems.

His action follows a Sept. 23 congressional hearing that examined the possibility of a binding, enforceable and verifiable U.S.-China agreement aimed at addressing catastrophic AI risks.

Khanna is seeking information from the three Chinese companies in several areas, including recursive self-improvement, artificial general intelligence, superintelligence, emergency shutdown mechanisms, model testing and international verification.

He asked whether the companies are developing, or intend to develop, AI systems capable of recursively improving themselves and what safeguards they would use to prevent humans from losing control.

The companies were also asked to provide details on their plans and timelines for developing artificial general intelligence or superintelligence.

Khanna is seeking information on whether the firms are developing emergency “kill switches” capable of disabling highly advanced AI systems if control is lost.

He also asked how models are tested before release, including whether testing takes place in air-gapped environments and whether systems are screened for capabilities that could pose severe risks, including the development of dangerous biological capabilities.

Khanna asked whether Chinese AI laboratories would participate in U.S.-China technical working groups aimed at developing methods to verify compliance with a future international AI agreement and whether they would allow independent inspections.

“Simply put, it is clear we cannot gamble the future of humanity on whether AI companies – including Chinese companies like yours – will responsibly self-regulate,” Khanna wrote.

Khanna had previously asked DeepSeek, Alibaba and Moonshot AI to join U.S. AI laboratories in a commitment to “pace the frontier” by slowing the development of increasingly advanced systems until safety measures could catch up.

The companies did not respond by the Sept. 22 deadline.

In his latest letters, Khanna criticized the companies, citing what he described as their involvement in human rights abuses, censorship, support for China’s military and intelligence services, and illicit distillation of U.S. AI models.

“With every refusal by your companies to participate in this effort to seek a verifiable global slowdown of AI development to benefit all of humanity, you only strengthen the case for a binding and enforceable AI treaty between our governments,” he wrote.

Khanna is also seeking an assessment from the U.S. intelligence community of the country’s ability to respond if control of an advanced AI system were lost.

He requested a separate intelligence assessment of Chinese President Xi Jinping’s and the Chinese Communist Party’s understanding of catastrophic AI risks, including misalignment and the possible use of advanced systems to facilitate biological, chemical or nuclear harm.

Khanna also urged U.S. intelligence agencies to strengthen their ability to monitor Chinese frontier AI laboratories and has proposed joint U.S.-China technical working groups involving leading scientists to develop verification technologies for a future international agreement.

“History will look back at this moment, the choices made by humanity as we potentially approach the Rubicon of superintelligence or AGI, and either celebrate our foresight or condemn our inaction,” Khanna wrote.

The debate comes as artificial intelligence becomes an increasingly prominent part of broader U.S.-China technology tensions, including disputes over advanced semiconductors and Chinese access to American technology and computing capabilities. (Source: IANS)