Anthropic co-founder and CEO Dario Amodei has cautioned U.S. lawmakers about the potential dangers of open-source artificial intelligence. Speaking to Congress, Amodei highlighted that once AI models with advanced capabilities are released as open-source, developers lose the ability to effectively supervise their use. This includes challenges in monitoring misuse, revoking access, and updating security measures, which could lead to significant and irreversible risks. Amodei emphasized that open-source AI models are more difficult to control compared to closed systems, posing a greater challenge for security governance. He warned that the lack of control over these models could lead to misuse, underscoring the need for careful consideration in the development and release of open-source AI technologies.