Former Google Engineer Convicted of Stealing AI Secrets for China

A 38-year-old former Google software engineer has been convicted of siphoning thousands of pages of the company’s most sensitive AI technology to benefit China, in one of the most serious tech espionage cases prosecuted by the Justice Department.

Linwei Ding, also known as Leon Ding, was found guilty on seven counts of economic espionage and seven counts of theft of trade secrets following an 11-day trial in San Francisco federal court, the U.S. Attorney’s Office for the Northern District of California said in a statement on Thursday.

In March 2024, federal agents arrested Ding at his Newark home, with then-U.S. Attorney Ismail Ramsey alleging he stole 500+ confidential AI files while secretly working with China-based companies to gain a competitive edge.

Ding now faces up to 15 years in prison for each economic espionage count and 10 years for each theft count, with a status conference scheduled for February 3.

The trial revealed how Ding systematically plundered Google’s AI crown jewels over nearly a year.

Between May 2022 and April 2023, while working as a Google software engineer with access to the company’s most confidential systems, he copied more than 2,000 pages of proprietary information to his personal Google Cloud account.

The trade secrets included detailed specifications for Google’s custom Tensor Processing Unit chips, Graphics Processing Unit systems, and the sophisticated software that orchestrates thousands of chips into supercomputers capable of training cutting-edge AI models.

While stealing Google’s AI secrets, Ding was also building his own China-based AI venture, discussing a CTO role with a Chinese startup in mid-2022 before founding and leading his own AI company as CEO by early 2023.

In investor pitches, Ding claimed he could replicate Google’s technology, promoted China’s state-backed AI priorities, and applied to a Shanghai government talent program in late 2023, pledging to help China reach “international-level” computing infrastructure.

“The jury delivered a clear message today that the theft of this valuable technology will not go unpunished,” U.S. Attorney Craig H. Missakian said in the statement.

Industrial espionage and AI

The conviction comes as Chinese exploitation of American AI technology has become a major concern for U.S. officials.

“AI models are already geopolitically sensitive as we are deep into the AI arms race, and the nation that leads this race will gain an advantage similar to those won by the U.S. and the Soviet Union in the nuclear arms race,” Kadan Stadelmann, Chief Technology Officer at Komodo Platform, told Decrypt.

“AI espionage is already running rampant, because there is so much at stake with this technology,” he added, urging AI startup founders to thoroughly vet engineers through comprehensive background checks.

In November, congressional investigators summoned Anthropic CEO Dario Amodei to explain how Chinese state actors used the company’s Claude Code tool to launch what Anthropic described as the first large-scale cyber operation largely automated by AI.

Last month, major U.S. tech companies formed the Agentic AI Foundation under the Linux Foundation, partly in response to China overtaking the U.S. in open-source AI downloads.