White House to other government agencies, tech giants in memo: Explore measures to stop China from stealing ‘American AI’

1 hour ago 6
ARTICLE AD BOX

 Explore measures to stop China from stealing ‘American AI’

The White House has issued an urgent directive to government agencies, accusing China of running a massive, state-sponsored campaign to steal the American artificial intelligence (AI) technology.

In a sharply worded memo released on Thursday (April 23), the administration claimed that foreign entities – primarily based in China – are engaged in “industrial-scale” efforts to copy and dismantle the United States’ most advanced AI systems.The memo, written by Michael Kratsios, Director of the White House Office of Science and Technology Policy, warns that these actors are using a technique known as “distillation” – a process involves studying a top-tier US AI model to create an imitation version that mimics its power without needing the original research and development.

Read full memo sent by Michael Kratsios

EMORANDUM FOR THE HEADS OF EXECUTIVE DEPARTMENTS AND AGENCIESFROM: MICHAEL J. KRATSIOSSUBJECT: ASSISTANT TO THE PRESIDENT FOR SCIENCE AND TECHNOLOGY DIRECTOR, OFFICE OF SCIENCE AND TECHNOLOGY POLICYAdversarial Distillation of American AI ModelsThe United States leads the world in artificial intelligence (AI) technologies. That lead reflects decades of foundational research, bold entrepreneurial risk-taking, and hundreds of billions of dollars in annual private investment. American AI leadership drives economic growth, strengthens national security, and advances the frontiers of science, medicine, and human knowledge. The breakthroughs emerging from American industry raise living standards, expand opportunity, and improve lives around the world.However, the United States government has information indicating that foreign entities, principally based in China, are engaged in deliberate, industrial-scale campaigns to distill U.S. frontier AI systems. Leveraging tens of thousands of proxy accounts to evade detection and using jailbreaking techniques to expose proprietary information, these coordinated campaigns systematically extract capabilities from American AI models, exploiting American expertise and innovation.Models developed from surreptitious, unauthorized distillation campaigns like this do not replicate the full performance of the original. They do, however, enable foreign actors to release products that appear to perform comparably on select benchmarks at a fraction of the cost. These distillation campaigns also allow those actors to deliberately strip security protocols from the resulting models and undo mechanisms that ensure those AI models are ideologically neutral and truth-seeking.The United States is committed to the free and fair development of AI technologies across a competitive ecosystem, from leading frontier models to highly-tuned applied systems, and from open-source frameworks to open-weight models. AI distillation, when legitimately used to produce smaller, lighter-weight models from more advanced systems, is a vital part of that ecosystem. Industrial distillation activities that aim to systematically undermine American research and development and access proprietary information, however, are unacceptable.To address this threat, the Trump Administration will:1. Share information with U.S. AI companies concerning attempts by foreign actors to conduct unauthorized, industrial-scale distillation, including the tactics employed and actors involved.2. Enable the private sector to better coordinate against such attacks.3. Work together with private industry to develop best practices to identify, mitigate, and remediate industrial-scale distillation activities and build strong defenses against such activities.4. Explore a range of measures to hold foreign actors accountable for industrial-scale distillation campaigns. There is nothing innovative about systematically extracting and copying the innovations of American industry, and there is nothing open about supposedly open models that are derived from acts of malicious exploitation.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.Consistent with America's AI Action Plan, the United States will continue to foster a vibrant open-source ecosystem built on firm foundations, support American industry in making frontier AI broadly accessible to users worldwide, and safeguard the free and fair market competition that enables the broad and beneficial diffusion of these technologies.

Read Entire Article