Unproven Nvidia Espionage Claims Revive a U.S. Security Debate

By Olivia Hart

U.S. intelligence officials have raised concerns about a possible Chinese intelligence effort involving Nvidia, according to reports in The New York Times and elsewhere; the allegations remain unproven. Nvidia says it has no reason to believe company secrets or U.S. national security were compromised, and Beijing has rejected the claims. The latest allegations have prompted investigations, denials and renewed expressions of alarm.

What do the Nvidia allegations establish?

Susan Katz Keating, the publisher and editor-in-chief of Soldier of Fortune, describes Nvidia as one of America’s most important technology companies and cites The New York Times and other reports in describing the concerns. Her account does not identify an alleged intelligence asset, a method of collection or supporting evidence, which leaves readers unable to test the warning against an identified mechanism.

“The so-called Chinese spy narrative is purely a groundless accusation, slander, and a smear,” Chinese Foreign Ministry spokesperson Guo Jiakun told reporters in Beijing on Monday. The statement records Beijing’s rejection but does not resolve what U.S. intelligence officials allege or what evidence may emerge.

Public relations firms, meanwhile, are offering interviews to Soldier of Fortune on “honeypot” operations. On the record presented, the immediate issue remains credibility: the account labels the allegations unproven, and both Nvidia and Beijing reject them.

Why is the reaction familiar?

Keating frames the central question as why the possibility of another Chinese intelligence operation still astonishes U.S. officials, even though the pattern is familiar. Her earlier essay, Same Spy, Different Day: The Recurring Shock of Chinese Espionage, called that response a dangerous case of collective amnesia.

Robert Crowley, whom Keating describes as her old friend, served as a former CIA assistant director for clandestine operations and was an expert on the Soviet KGB. Keating recalled that he believed American officials failed to recognize the scale of Chinese intelligence operations even after repeated warnings and damaging security breaches.

“China viewed America as a country that left its gold lying in the streets, Beijing simply gathered up the unguarded treasure,” Crowley told Keating. Keating says he also feared that a well-placed intelligence asset could cause serious damage before anyone recognized the danger. The warning was less about a new threat than a failure to remember earlier ones.

What does the historical record show?

Larry Wu-Tai Chin spent decades passing American intelligence secrets to China while working inside the U.S. intelligence establishment. His espionage remained undetected until information from a defector exposed him. Former CIA Director William Colby told Keating that Chin’s betrayal came as a total shock.

American investigators later uncovered Chinese efforts to acquire sensitive weapons technology, nuclear secrets and military equipment. A 1999 congressional investigation detailed China’s acquisition of information about U.S. nuclear warhead designs.

Former CIA officer Alexander Yuk Ching Ma was later arrested for conspiring to provide classified information to Chinese intelligence. He pleaded guilty in 2024 and was sentenced to 10 years in federal prison. Keating uses Ma’s case to show how trusted positions can channel sensitive information without immediate detection.

Keating says the methods varied: some operatives worked from positions of trust inside government agencies, while others pursued military technology, scientific research or commercial information. Each exposure produced renewed warnings followed by assurances that American institutions understood the danger. As Keating presents it, that record points to varied insider access rather than one fixed playbook.

Why do commercial relationships complicate espionage?

Keating argues that advanced semiconductors powering artificial intelligence now rank among the most valuable targets for foreign intelligence services. She says this makes the U.S. artificial-intelligence industry an obvious area of interest regardless of what the current allegations establish.

Keating says the commercial boundary is difficult to police because an executive can meet thousands of customers, suppliers and business partners without knowing which encounters might attract a foreign intelligence service. She says the problem extends beyond Nvidia and beyond semiconductors, but her account does not claim that any named commercial contact played a role in the current case.

Crowley understood that espionage did not always require breaking into a vault. Sometimes information became accessible through ordinary relationships cultivated over time, making a routine encounter harder to identify. The practical difficulty is detecting the hidden purpose before the information changes hands.

What can the current case establish?

The latest allegations may eventually produce evidence of an intelligence operation or leave important questions unresolved. Keating’s broader argument does not depend on that outcome: she says China has pursued American military, scientific and technological secrets for decades, while successive generations of U.S. officials have greeted each exposure with fresh astonishment.

Her remedy for lawmakers is blunt: keep a copy of the historical record on their desks rather than rely on another round of shocked expressions. For now, the test is whether investigators produce evidence tying a Chinese intelligence operation to Nvidia; until they do, the established history warrants caution without replacing proof with assumption.