Dario Amodei, Anthropic CEO: "the most dangerous model could be one trained in secret and delivered only to the Chinese army"

Anthropic CEO Dario Amodei warns that the most dangerous AI model could be secretly trained and handed solely to the Chinese military. He points out that the rivalry between the West—represented by Anthropic, OpenAI and xAI—and China—home to Moonshot AI, Baidu and Alibaba—is increasingly focused on model openness. Amodei expresses concern over open‑model strategies as a competitive threat.

Summary written by Kernelia from the original article by Xataka. The story and its rights belong to its author.