Find notable cyber news and cases, enriched with sources, timelines, and signals.
Home Watchlists Beta Browse News Feed Stats About Contact

Microsoft launches agent guardrails, identities, and Security Copilot updates for agentic AI

Security Tool/Service
First reported
Last updated
Happening score
H score 11
1 unique sources, 1 articles

Summary

Hide ▲

Microsoft rolled out new agentic AI security controls at RSAC Conference, adding preview guardrails in Microsoft Foundry, agent identities in Entra ID, and updated Security Copilot capabilities. The changes are meant to help organizations track, limit, and investigate AI agents as autonomous systems expand the attack surface.

Related Happenings

Microsoft Teams admin policy adds approval-based control for third-party bots

Security Tool/Service
H score11 First: 30.06.2026 13:52 Last: 30.06.2026 13:52 Sources 1

About this happening: Microsoft Teams introduced an **admin policy** that lets organizers prevent **third-party bots** from joining meetings without approval. The control improves **visibility** over e...

Open Happening

Microsoft Teams third-party bot approval controls for meeting social-engineering risk

Defensive Guidance
H score11 First: 30.06.2026 13:52 Last: 30.06.2026 13:52 Sources 1

About this happening: **Microsoft Teams** has added admin controls that block **third-party bots** without approval, reducing **meeting social-engineering risk** across managed tenants. The policy impr...

Open Happening

Microsoft open-sources RAMPART and Clarity for AI agent security testing and design review

Security Tool/Service
H score10 First: 20.05.2026 20:06 Last: 20.05.2026 20:06 Sources 1

About this happening: **Microsoft** open-sourced **RAMPART** and **Clarity**, adding **AI agent security testing** and **design-time reasoning** capabilities that help developers catch risks before dep...

Open Happening

KongTuke Microsoft Teams initial access campaign

Campaign
H score42 First: 14.05.2026 15:12 Last: 14.05.2026 15:12 Sources 1

About this happening: The **KongTuke** campaign now uses **Microsoft Teams** social engineering to gain persistent access to **corporate networks**, shortening initial compromise to **under five minute...

Open Happening

Microsoft Teams remote assistance abuse mitigation

Advisory/Mitigation
H score15 First: 20.04.2026 18:11 Last: 20.04.2026 18:11 Sources 1

About this happening: **Microsoft** issued mitigation guidance to curb **Teams-adjacent remote assistance abuse**, warning that external contacts should be treated as untrusted and that **remote assist...

Open Happening

Timeline

  1. 24.03.2026 14:28 2 articles · 3mo ago

    Microsoft launches agent identities and guardrails for agentic AI

    Mitigation Patch Update

    Microsoft launched preview guardrails in Microsoft Foundry and special agent identities in Azure AI Foundry managed by Microsoft Entra ID, giving organizations a registry for AI agents plus controls for permissions, logging, and limiting autonomous agent behavior at RSAC Conference.

    Show sources
    Open in new tab
  2. 24.03.2026 14:28 1 articles · 3mo ago

    Microsoft discloses Security Copilot and Zero Trust updates for autonomous agents

    Initial Disclosure

    Microsoft disclosed new agentic AI security controls at RSAC Conference, including agent identities in Azure AI Foundry, expanded guardrails for models and agents, and Security Copilot updates that add a Security Triage Agent, a Security Analyst agent, and deeper investigation workflows using Microsoft Defender and Sentinel telemetry.

    Show sources
    Open in new tab