Microsoft admits 'frontier models can't be trusted', unveils new security tools to treat AI agents as insider threats

T2Covered by 1 source · first reported by TechRadar / Tom's Guide / PCWorld / Pocket-lint / Stuff at 20 Oct, 11:44 UTC

Say hi to your new Guardian Agent, keeping you safe from Agentic Insiders.

consumer

Sources

• T2Microsoft admits 'frontier models can't be trusted', unveils new security tools to treat AI agents as insider threatsTechRadar / Tom's Guide / PCWorld / Pocket-lint / Stuff20 Oct, 11:44