Ex-Google Engineer Convicted Of Stealing AI Secrets For Chinese Companies

In a landmark case underscoring the escalating global race for artificial intelligence dominance, a former software engineer for a leading US technology firm, Linwei Ding, has been convicted of stealing crucial AI trade secrets. A federal jury in San Francisco delivered the guilty verdict on Thursday, finding Ding culpable of diverting highly sensitive information to two Chinese companies while secretly in their employ.

The 38-year-old Chinese national was found guilty on a total of fourteen counts following an intensive eleven-day trial. The charges included seven counts of economic espionage and seven counts of theft of trade secrets, reflecting the gravity and scale of his illicit activities. Prosecutors meticulously presented evidence detailing how Ding absconded with thousands of pages of confidential data, directly compromising the advancements of his former employer in the critical field of artificial intelligence.

The stolen information encompassed the very backbone of the company’s advanced AI capabilities: proprietary designs related to its hardware infrastructure and the sophisticated software platform underpinning its supercomputing data centers. These powerful systems are essential for training the large AI models that are at the forefront of modern technological innovation. The blueprints for these next-generation chips were not merely technical specifications; they represented a strategic advantage, designed to bolster the company's competitive edge against rivals in cloud computing and lessen its dependency on external chip suppliers.

According to court documents, Ding commenced his employment with the US tech giant in May 2019. The elaborate scheme of intellectual property theft began approximately three years later, coinciding with a period when Ding was actively being courted by nascent Chinese technology ventures. This timeline suggests a premeditated effort to leverage his insider access for personal gain and to benefit foreign entities, raising serious questions about corporate espionage in the high-stakes world of AI development.

The ramifications of such convictions are severe. Each count of economic espionage carries a maximum potential sentence of 15 years in federal prison and a hefty fine of up to $5 million. Similarly, each count of trade secret theft could lead to a maximum of 10 years incarceration and a $250,000 fine. While Ding was slated for a status conference in early February, the verdict itself sends a clear message regarding the stringent legal protections afforded to intellectual property, particularly in sectors vital to national economic security.

This case was a key component of the interagency Disruptive Technology Strike Force, an initiative established in 2023 by the US administration to counteract threats posed by adversaries misusing disruptive technologies and stealing critical intellectual property. The successful prosecution of Ding highlights the efficacy of this focused effort to safeguard American innovation and prevent sensitive technological advancements from being illicitly transferred abroad.

The company confirmed its full cooperation with law enforcement agencies throughout the investigation and legal proceedings, reaffirming its commitment to protecting its proprietary technologies. The conviction serves as a stark reminder to tech professionals globally about the severe consequences of betraying trust and engaging in corporate espionage, particularly when it involves highly strategic and transformative technologies like artificial intelligence.