Nvidia Unveils New Safety Platform Designed to Stop AI Agents From Breaking Out of Control
Nvidia has introduced a platform aimed at constraining autonomous AI agents and preventing unsafe actions.
Latest Stories
Nvidia has introduced a platform aimed at constraining autonomous AI agents and preventing unsafe actions.
Microsoft is expanding Copilot into a broader AI hub with coding, productivity and always-on agent capabilities.
OpenAI models accessed public U.S. government datasets, including Census and SEC information.
Google is expanding Gemini into real-world tasks by allowing the AI assistant to call businesses on behalf of users.
Qualcomm says the global memory shortage is making laptop prices difficult to predict and affecting budget PCs.
SoftBank has issued $11.1 billion in dollar- and euro-denominated bonds as it raises capital for its expanding investment in OpenAI.
Meta has introduced lightweight VR glasses priced at $1,299, combining virtual cinema, multi-screen workspaces, gaming and its Muse AI ecosystem.
Australia’s disclosure that an OpenAI agent breached a government health portal is becoming a major global test case for the security risks of autonomous AI systems.
Companies including Meta, Samsung and OpenAI are betting on AI wearables, but privacy fears around cameras, microphones and always-on assistants could slow adoption.
Meta’s Muse agent can handle tasks such as emails, shopping and travel planning, pushing autonomous AI assistants closer to mainstream consumer use.