U.S. Accuses China of Massive AI Theft Campaigns Ahead of Trump’s Beijing Visit

U.S. Government Accuses China of Industrial-Scale AI Theft

The Trump administration on Thursday accused China-backed actors of running "deliberate, industrial-scale campaigns" to steal and replicate American frontier AI models. The accusation intensifies the U.S.-China AI rivalry and could complicate President Trump’s planned visit to Beijing next month.

White House Memo Details Distillation Attacks

Michael Kratsios, director of the White House Office of Science and Technology Policy, sent a memo to federal agency heads on Thursday. The memo accused mostly China-based actors of using proxy accounts to evade detection and "jailbreak" AI models to expose proprietary information and extract capabilities.

These distillation attacks involve querying proprietary models—such as Claude or Gemini—millions of times via APIs to build datasets that replicate how the systems behave.

Kratsios stated that these campaigns allow foreign actors to release AI models that appear to match U.S. capabilities at a fraction of the cost. He also warned that such tactics can remove guardrails designed to keep outputs "ideologically neutral and truth-seeking."

Upcoming Trump Visit to Beijing Adds Stakes

The warning comes as President Trump prepares for a highly anticipated trip to Beijing next month. During the visit, Trump is expected to push for economic concessions and reset parts of the U.S.-China relationship.

Major AI Firms Report Distillation Attacks from China

Earlier this year, OpenAI and Anthropic reported that China-based firms—including DeepSeek, Moonshot AI, and MiniMax—were behind large-scale distillation attacks on their models.

Broader Context: U.S.-China AI Espionage Concerns

The U.S. has long accused China of stealing intellectual property from American companies as part of broader cyber espionage efforts. In 2024, the Justice Department indicted a former Google software engineer for stealing AI trade secrets and sharing them with two Chinese companies.

White House Warns of Risks in Distilled AI Models

Kratsios argued that the capabilities of these stolen and replicated models may not hold up over time. He stated:

"As methods to detect and mitigate industrial-scale distillation grow more sophisticated, foreign entities who build their AI capabilities on such fragile foundations should have little confidence in the integrity and reliability of the models they produce."

Next Steps: U.S. Plans to Share Intelligence with AI Firms

Kratsios said the Trump administration plans to share intelligence with U.S. AI companies about these campaigns, including the tactics used. The goal is to help the private sector develop stronger defenses against such attacks.