Dario Amodei: Supports open-weight AI but fears China's growing capabilities
Anthropic CEO Dario Amodei has clarified his stance on open-weight AI models and China's rapidly advancing AI capabilities. While he does not oppose open-weight models outright, he expresses serious concern about how China could leverage such technology. Amodei emphasizes the need for responsible AI development amid growing geopolitical tensions.
Dario Amodei, the founder and CEO of AI safety company Anthropic, has stepped forward to clarify his position on open-weight AI models after facing criticism suggesting he was opposed to them. In a conversation with TechCrunch, Amodei made clear that his concerns are not rooted in opposition to openness itself, but rather in the geopolitical risks that come with the uncontrolled spread of powerful AI technology.
China stands at the center of Amodei's concerns. He argues that Chinese actors could exploit openly available AI models to rapidly develop capabilities that may be applied in military or intelligence contexts. This, in his view, represents one of the most urgent security challenges currently facing the AI industry and the broader tech ecosystem.
At the same time, Amodei acknowledges the significant benefits that open-weight models bring to the table. They democratize access to AI, enabling researchers, startups, and developers around the world to innovate without relying solely on proprietary systems. He is careful to distance himself from being labeled an opponent of the open-source AI movement, framing his stance instead as a call for thoughtful risk management.
Amodei's comments reflect a broader tension within the technology industry: how to balance the ideals of openness and innovation with the realities of national security and geopolitical competition. As AI becomes increasingly central to global power dynamics, his perspective adds an important voice to the ongoing debate about who should govern AI development and how.