Anthropic limits AI model release over security concerns

AI firm Anthropic has restricted the release of its new model, Mythos, citing its advanced capabilities in finding software security exploits. The model will be shared with select large organizations and critical infrastructure providers, including AWS and JPMorgan Chase, to proactively address potential vulnerabilities. This controlled release strategy aims to allow these entities to patch weaknesses before malicious actors can exploit them using advanced AI. However, some experts suggest this move also serves to protect Anthropic's business model by limiting the ability of smaller competitors to use distillation techniques to replicate their cutting-edge models, thereby preserving the value of enterprise contracts.