Nvidia has launched a new security platform designed to keep artificial intelligence agents from going rogue, addressing increasing concerns about the safety and management of AI technologies. This open-source initiative, dubbed the Open Agent Safety Platform, aims to provide tools that can contain rogue AI behavior within milliseconds. The platform has been developed in collaboration with over 100 partners and includes software and hardware components to detect and manage AI agents effectively. Nvidia claims that this new system could have prevented incidents like the Hugging Face hack, showcasing its potential to reinforce AI security in various applications. Additionally, Nvidia has announced a significant stock buyback plan alongside this launch, indicating its confidence in the future demand for its products amid evolving regulatory landscapes and industry debates surrounding AI safety.