Anthropic’s CEO, Dario Amodei, has raised serious concerns about the potential for foreign espionage targeting top U.S. artificial intelligence firms—particularly from China. He believes that highly valuable “algorithmic secrets” worth millions of dollars could be falling into the hands of foreign intelligence operatives, and he is urging the U.S. government to take action.
During an event at the Council on Foreign Relations on Monday, Amodei pointed out that China has a long history of engaging in large-scale industrial espionage. Given the immense value of proprietary AI advancements, he is convinced that companies like Anthropic are prime targets for such operations.
“Some of these algorithmic breakthroughs represent $100 million in value but can be reduced to just a few lines of code,” Amodei explained. “It’s highly likely that there are efforts to steal them—and they may already be succeeding.”
While he stopped short of specifying exactly what measures should be implemented, Amodei stressed the importance of increased U.S. government support in safeguarding the industry from such threats.
Although Anthropic declined to comment directly on Amodei’s statements when approached by TechCrunch, the company did reference its recent policy recommendations submitted to the White House’s Office of Science and Technology Policy (OSTP). In these recommendations, Anthropic advocates for a partnership between the federal government and leading AI firms to enhance security protocols at frontier AI research labs. This includes potential collaboration with U.S. intelligence agencies and their allies to strengthen cybersecurity measures.
Amodei’s position reflects his broader skepticism about China’s AI ambitions, particularly regarding its use for military and authoritarian applications. He has previously advocated for stringent U.S. export controls on AI chips bound for China. Additionally, he pointed out that DeepSeek, a Chinese AI model, performed the worst in a critical bioweapons safety test conducted by Anthropic.
His warnings, as outlined in his essay Machines of Loving Grace, emphasize the potential dangers of China harnessing AI for geopolitical dominance. However, some AI researchers argue that fostering cooperation between the U.S. and China, rather than intensifying restrictions, could help prevent an AI arms race that might lead to the development of uncontrollable artificial intelligence systems.
The debate over AI security and international collaboration continues to intensify, raising critical questions about the future of global AI governance and the risks of technological espionage.
