US Names Six Chinese AI Companies for Alleged Large‑Scale Copying of Frontier Models

The United States government announced Tuesday that it has identified six Chinese artificial‑intelligence firms it says have engaged in systematic efforts to replicate cutting‑edge U.S. AI models, a practice officials describe as an "industrial‑scale attack" on American technology leadership.

The designation was issued jointly by the National Security Agency and the Cybersecurity and Infrastructure Security Agency, which together asserted that the companies have been extracting core capabilities from leading U.S. models and re‑creating them at a fraction of the development cost.

According to the agencies, the alleged activity involves reverse‑engineering publicly available AI outputs, training large language models on those outputs, and then commercialising the results under their own brands. Officials contend that this approach could save the Chinese firms billions of dollars that would otherwise be required to fund original research and development.

U.S. officials did not disclose the names of the firms in the public release, but the move signals a sharpening of concerns in Washington about intellectual‑property theft and the rapid narrowing of the AI gap between the two nations. The United States has long warned that unchecked copying could erode its competitive edge in a sector increasingly tied to national security, economic growth, and geopolitical influence.

Chinese authorities have previously dismissed similar accusations as unfounded, arguing that their AI advancements stem from domestic innovation and open‑source collaboration. No official comment was immediately available from the companies named, but industry analysts expect they will deny wrongdoing and may seek diplomatic engagement to mitigate potential restrictions.

The announcement arrives amid a broader U.S. strategy to tighten controls on emerging technologies. Recent executive orders have expanded export‑control lists to cover advanced AI hardware and software, and the Department of Commerce is reportedly reviewing additional measures that could target entities involved in illicit technology transfer.

Experts say the naming could pave the way for sanctions, licensing restrictions, or other punitive actions that would limit the firms' ability to access U.S. cloud services, semiconductor components, or research collaborations. Such steps would aim to raise the cost of copying and encourage indigenous development that respects international IP norms.

As the AI race intensifies, the episode underscores the challenges of policing a field where open research, rapid diffusion, and cross‑border collaboration are commonplace. Observers note that the effectiveness of naming and shaming will depend on the ability of policymakers to enforce penalties without stifling legitimate innovation or escalating broader trade tensions.