Microsoft Brings Perception Agentic Security Platform to General Availability

Microsoft used its Ignite 2026 conference to announce the general availability of Perception, an agentic security platform that coordinates teams of specialized AI agents. The system, previously known as Project Perception, has been in public preview inside Microsoft Defender since August and now offers four generally available blue-team agents along with one red-team agent. Red agents simulate attacks, blue agents manage detection and investigation, and green agents apply fixes across Microsoft, AWS, and Google Cloud environments. The Alert Triage Agent is reported to handle roughly 80 percent of alerts in a typical Defender queue based on analysis of approximately 8 million alerts. Microsoft also previewed integrations with SIEM, SOAR, and IT service management tools while introducing capabilities to detect and control rogue AI agents such as Claude Code, OpenAI Codex, and GitHub Copilot.

SecurityWeekTranslated from English

Read at source

Microsoft used its Ignite 2026 conference on Tuesday to announce the general availability of Perception, a product designed for coordinating teams of AI agents to handle security tasks.

Unveiled in July as Project Perception, the system has been in public preview inside Microsoft Defender since August. Playbooks direct the agents, which are split into red, blue and green teams. Red agents play the attacker and probe for weaknesses; blue agents handle detection, alert triage, and investigations; and green agents fix problems.

Generally Available Agents

The four blue team agents that are now generally available cover alert triage, attack investigation, detection authoring and threat intelligence. According to Microsoft, the Alert Triage Agent can handle roughly 80% of alerts in a typical Defender queue, based on an analysis of approximately 8 million alerts across its security products.

The Recon Agent, one of the red team agents, is also generally available, and a Web App Pen Test Agent has entered preview. Two new green team playbooks work across Microsoft, AWS and Google Cloud. The agents behind them include Posture Prioritization, which is generally available, and Risk Reduction, which is in public preview.

The Risk Reduction Agent can recommend vulnerability shielding when a fix isn’t immediately available. This preview feature temporarily protects vulnerable assets until a permanent patch can be rolled out.

“Your team sets the guardrails and signs off on high-impact actions, and automated triggers launch playbooks when an incident meets the criteria you set,” Aleš Holeček, president and chief architect of Microsoft Security, wrote in a blog post.

Integrations and AI Agent Risk Controls

Microsoft is also previewing integrations that link Perception to SIEM, SOAR and IT service management tools via MCP and APIs. The company also addressed the risk posed by AI agents, which it views as a new kind of insider with broad access to data and tools.

In preview, Agent 365 and Entra Suite can detect 60 types of local AI agents, including Claude Code, OpenAI Codex and GitHub Copilot. Intune policies can keep supported agents on the device.

Meridian digestThe week's most consequential security stories, distilled into one briefing.See the digest

More in AI & Model Risk

Researchers Show Malicious Webpage Could Hijack Meta Muse AI Agent and Extract Private Conversations
China-Linked Actors Use Autonomous AI Agents to Discover Zero-Day Path Traversal in MONAI Label and Achieve Root Access
AI & Model Risk

China-Linked Actors Use Autonomous AI Agents to Discover Zero-Day Path Traversal in MONAI Label and Achieve Root Access

Suspected China-linked hackers deployed autonomous AI agents powered by the open-source Cairn framework to identify and exploit a previously unknown path traversal vulnerability in MONAI Label, an open-source medical imaging server used by research teams. The flaw, tracked as CVE-2026-133968 and rated 9.8, affects versions 0.8.5 and earlier by allowing unauthenticated attackers to write files outside the designated storage directory when authentication is disabled by default. Amazon Threat Intelligence reconstructed the campaign after discovering an exposed Cairn control panel that revealed the agents' prompts, task settings, and execution history spanning roughly 48 hours in September. Within about 11 minutes the agents mapped the application and achieved arbitrary file writes; in a follow-up session they planted a malicious Python startup file that was loaded during a training job, granting root access on a university server. Maintainers merged a source-code fix on October 19 that enforces path resolution inside approved folders, yet no patched release has been published and the latest available version remains 0.8.5 from 2024. Amazon attributes the activity to a China-linked operator with moderate confidence based on Simplified Chinese prompts and China Standard Time settings, though no specific group or government sponsorship was confirmed.

The Hacker NewsNov 21, 2026Translated from English

Chinese-Linked Actors Use Cairn AI Platform to Discover and Exploit MONAI Label Zero-Day