Article

US Agencies Accuse Chinese AI Firms of Industrial-Scale IP Theft

The United States intelligence community and federal law enforcement agencies have jointly accused major Chinese artificial intelligence firms—including DeepSeek and Alibaba—of conducting what they describe as "industrial-scale" theft of American AI trade secrets. The allegations mark one of the most aggressive public confrontations yet over the origins of cutting-edge AI models emerging from China.

The Core Allegations

U.S. agencies allege that the companies systematically copied proprietary American AI models through a coordinated effort. The characterization of the activity as "industrial-scale" signals that investigators believe the copying extended beyond isolated incidents, describing a deliberate, high-volume operation rather than incremental or accidental replication.

The accusations rest on intelligence assessments and evidence gathered through law enforcement channels. Agencies involved in the joint statement have framed the alleged activity as a deliberate strategy to close the technological gap between Chinese and American AI capabilities, leveraging copied American innovations rather than independent research trajectories.

DeepSeek and Alibaba Named

Among the firms named in the allegations, DeepSeek has drawn particular attention. The company gained global notice in early 2025 when it released a high-performing large language model at a fraction of the typical development cost, quickly disrupting global AI market assumptions. The speed and cost efficiency of DeepSeek's advances had already prompted scrutiny in Western policy circles, with some analysts questioning whether the results were consistent with independent development timelines.

Alibaba, a long-established technology conglomerate with its own cloud and AI divisions, is also cited in the allegations. The company has been investing heavily in AI capabilities and has promoted its latest models as competitive with leading American products.

Trade Secret Framework

The legal framing matters. By labeling the alleged activity as theft of trade secrets—rather than copyright infringement or patent violation—U.S. authorities are invoking the Defend Trade Secrets Act, a federal statute that carries both civil and criminal penalties. Trade secret claims are notably broader than patent claims, as they do not require public disclosure of the method or technology in question.

The "malicious" characterization used in agency communications signals potential criminal referrals, though no formal charges against specific individuals or entities were detailed in the initial statements.

Geopolitical Context

The accusations arrive amid escalating technology competition between Washington and Beijing. The U.S. has progressively tightened export controls on advanced semiconductors and AI-related hardware to China, seeking to limit the infrastructure available for frontier AI development. The allegations suggest that, in the view of U.S. agencies, software-level acquisition of American AI intellectual property represents a parallel and harder-to-control avenue of technology transfer.

China has consistently denied state involvement in technology theft and rejects characterizations of its AI industry as dependent on foreign innovation. Chinese officials have pointed to significant domestic investment in AI research and a rapidly growing talent pool as the foundation for recent advances.

Industry and Market Reaction

The accusations have reverberated through markets and policy discussions. AI-related stocks saw volatility following the initial reports, and technology policy analysts expect the allegations to intensify calls for stronger enforcement of existing IP frameworks and possibly new restrictions on AI-related partnerships between American firms and Chinese entities.

The incident is likely to complicate ongoing discussions about international AI governance, model safety standards, and export control regimes, as it introduces a new layer of distrust into multilateral technology negotiations.

What Comes Next

U.S. agencies have indicated that the investigation is active, with further evidence expected to be presented. Whether the allegations result in formal charges, diplomatic fallout, or new regulatory measures will depend on the strength of classified evidence and the political calculus on both sides. The case also raises procedural questions about how trade secret theft claims involving frontier AI models would be adjudicated, given the technical complexity and the international nature of the entities involved.

Sources