AI TRENDS | Anthropic CEO Dario Amodei Says Open-Source AI Is Entering a Dangerous Path

Anthropic co-founder and CEO Dario Amodei told U.S. lawmakers that the development of open-source AI is entering what he described as a “very dangerous path,” warning that releasing highly capable AI models as open source can reduce oversight of how they are used. According to Odaily, Amodei said that once powerful AI models are published in open-source form, developers may lose effective control over their deployment, including the ability to monitor misuse, revoke access, or dynamically update safety protections. He added that compared with closed-model systems, fully open models are more difficult to manage under ongoing safety governance and could create risks of misuse that are hard to reverse.