Dario Amodei, CEO of AI startup Anthropic, has a warning. He believes spies, likely from China, are after the U.S.’s top AI secrets. These secrets, worth $100 million, may exist in just a few lines of code.
Speaking at a Council on Foreign Relations event, Amodei stressed the urgency. China, he said, is known for large-scale industrial espionage. He suspects AI companies, including Anthropic, are already targets.
“Many of these algorithmic secrets, there are $100 million secrets that are a few lines of code,” Amodei stated. “And, you know, I’m sure that there are folks trying to steal them, and they may be succeeding.”
A Call for Government Action
Amodei didn’t just point out the threat, he also urged action. He called on the U.S. government to do more. However, he didn’t specify exactly what steps should be taken. Anthropic declined to elaborate on his remarks.
Instead, the company referred to its recent recommendations to the White House’s Office of Science and Technology Policy (OSTP). In those recommendations, Anthropic suggested the federal government work closely with AI companies. This collaboration, they argue, should include U.S. intelligence agencies.
Why AI Secrets Matter
AI breakthroughs depend on unique algorithms and training methods. These take years, and millions of dollars, to perfect. If stolen, they could be replicated or improved at a fraction of the cost.
Leading AI companies, like OpenAI, Google DeepMind, and Anthropic, invest heavily in research. If a rival gains access to their work, they could skip the hard part and reap the benefits.
China’s AI Ambitions and U.S. Export Controls
Amodei has long voiced concerns about China’s AI development. He supports strict export controls on AI chips to China. He fears the technology could be used for surveillance or military purposes.
Not everyone agrees. Some AI researchers warn that restricting China too much could backfire. They argue that cutting off collaboration might lead to an AI arms race. In their view, cooperation could make AI safer and more controllable.
AI and National Security
The fight over AI is also about national security. Advanced AI could be used in many ways, including:
- Cyberwarfare: AI-driven attacks on infrastructure and finance systems.
- Biotechnology: AI models analyzing chemical and biological threats.
- Surveillance: Facial recognition and mass data tracking.
These risks have caught the attention of U.S. policymakers. The Biden administration has already taken steps to impose restrictions on China’s access to advanced AI chips. Meanwhile, discussions on AI regulations continue in Washington.
