Svmuu News: Dario Amodei, co-founder and CEO of Anthropic, told lawmakers in the U.S. Congress that the development of open-source AI is heading down a “very dangerous path.” Once powerful AI models are released as open-source, developers will lose the ability to effectively oversee how the models are used—including the inability to monitor misuse, revoke access, or dynamically update security measures—thereby increasing potential risks. Dario Amodei noted that, compared to closed-source model systems, fully open models are more difficult to subject to ongoing security governance, potentially creating irreversible risks of misuse.