US Senator Warns of China-Linked Insider Threats to Military AI Systems

Banks, who represent Indiana, raised the concerns in a letter to Defence Secretary Pete Hegseth and Office of Management and Budget Director Russell Vought, calling for closer coordination between the government and private AI developers.

The senator said the growing deployment of advanced AI in military operations was creating new vulnerabilities that existing corporate security programmes might not be sufficient to address.

He specifically pointed to potential threats from China, arguing that Beijing was seeking access to advanced US technologies.

"As the Department relies more on these companies, insider threats in the frontier SI industry are becoming a direct risk to military operations," Banks wrote, using the administration's term for superintelligence.

Banks pointed to agreements announced in May with companies including OpenAI and Google that were aimed at accelerating the deployment of advanced AI capabilities across the US military. He said such partnerships were likely to become increasingly important as AI systems gained capabilities relevant to cyber operations and military planning.

At the same time, he warned that the technology could become a target for foreign intelligence operations. According to Banks, an insider with sufficient access could potentially expose confidential information about military applications, steal AI model weights or introduce software vulnerabilities that could compromise defence systems.

The senator cited the January conviction of a former Google engineer on economic espionage charges involving the theft of confidential information related to the company's AI technology for the benefit of businesses based in China. He presented the case as an example of the risks posed by insider access to sensitive technological assets.

Banks acknowledged that major AI companies already operate internal security programmes but argued that the private sector could not address the threat on its own.

He called for continued coordination between the Pentagon and frontier AI developers to assess and mitigate insider risks.

The senator asked the Pentagon to obtain information from AI companies about their counterintelligence measures, the technology and data they consider most sensitive and the controls governing employee access.

He also sought an assessment of potential weaknesses in existing corporate security systems and the additional resources needed to address them.

Banks requested a staff-level briefing covering major national security risks, the Pentagon offices responsible for tackling insider threats and any additional funding or personnel required for counterintelligence efforts.

He also urged Congress, the Pentagon and the Office of Management and Budget to coordinate on providing adequate resources to strengthen counterintelligence and threat-reduction measures.

The letter did not identify any new espionage incident involving companies participating in US military AI partnerships, nor did it announce a new law or mandatory security requirement.

Instead, Banks called for greater government oversight and information sharing as the US military increases its reliance on private technology companies for advanced AI capabilities.

His concerns come amid intensifying competition between the United States and China over advanced computing, artificial intelligence and technologies that could influence future military capabilities.