Nvidia Introduces Security Platform to Prevent AI Agents From Going Rogue
Nvidia has announced a new security platform aimed at preventing AI agents from going out of control. While over a hundred institutions have adopted the platform, OpenAI and some major tech giants were not among the supporters.
Nvidia has introduced the Open Agent Safety Platform, developed to prevent artificial intelligence agents from exceeding boundaries and making unauthorized intrusions into external organizations. Combining open-source software and hardware monitoring layers, the system has generated widespread resonance in the industry.
The Purpose of the Open Agent Safety Platform
Nvidia has announced its new security platform to prevent artificial intelligence agents from getting out of control. This system aims to stop AI agents from exceeding their boundaries and making unauthorized intrusions into external organizations.
Software and Hardware Components
The platform includes an open-source software component called OpenShell, which allows developers to officially verify an agent's authority. Additionally, the system features a hardware-based security layer named Sentry that independently monitors and quarantines suspicious behavior.
Supporting Institutions and Companies
Nvidia announced during the launch that over a hundred organizations, including firms such as Microsoft, Perplexity, Accenture, JPMorgan Chase, Arm, and Intel, are using the platform.
The Status of OpenAI and Other Companies
OpenAI was not included in the list of supporters for Nvidia's platform. Although an OpenAI spokesperson stated that they support Nvidia's work and are collaborating on OpenShell, the company has not yet officially joined the platform.
Share Buyback Program
Nvidia also announced that its board of directors approved increasing the share buyback program by $150 billion, bringing the total to $235 billion.