Back to updates

Ai-Agents - 1

2026 (12)

May (5)

security: Defense in depth for autonomous AI agents

Microsoft Security Blog

2026-05-14

Microsoft says autonomous AI agents require updated defense-in-depth controls focused on application-layer design, identity, and human oversight. The post frames these as the main security areas to address as agent autonomy increases.

security: When prompts become shells: RCE vulnerabilities in AI agent frameworks

Microsoft Security Blog

2026-05-07

Microsoft says new research found prompt injection issues in AI agent frameworks can lead to remote code execution. The post outlines affected areas and recommends securing agents against these flaws.

dotnet: Durable Workflows in the Microsoft Agent Framework

.NET

2026-05-06

Microsoft Agent Framework now supports durable AI agent workflows using the Durable Task runtime. The post shows a progression from in-process console apps to parallel agent execution and Azure Functions hosting for serverless runs.

dotnet: Microsoft Agent Framework – Building Blocks for AI Part 3

.NET

2026-05-04

Part 3 of the Microsoft Agent Framework series shows how to build .NET AI agents using tools, multi-turn conversations, memory, and graph-based workflows. It ties together the earlier parts into the framework’s main building blocks for agent development.

security: Microsoft Agent 365, now generally available, expands capabilities and integrations

Microsoft Security Blog

2026-05-01

Microsoft announced general availability of Agent 365 and previewed new features for discovering and managing shadow AI agents, including local agents such as OpenClaw and Claude Code.

April (4)

ai: Introducing Toolboxes in Foundry

Microsoft AI Foundry Blog

2026-04-22

Microsoft Foundry introduces Toolboxes in public preview as a way to curate, configure, and reuse tools across AI agents. It is intended to reduce duplicated tool wiring, authentication, credentials, and integration code across frameworks and runtimes.

ai: From Local to Production: The Complete Developer Journey for Building, Composing, and Deploying AI Agents

Microsoft AI Foundry Blog

2026-04-22

Microsoft announces v1.0 of Microsoft Agent Framework and general availability of the Foundry Toolkit for Visual Studio Code. The post frames these releases as part of a workflow for building, composing, and deploying AI agents from local development to production.

security: The agentic SOC—Rethinking SecOps for the next decade

Microsoft Security Blog

2026-04-09

Microsoft describes a future SOC model where autonomous defense handles machine-speed response, agents provide context and coordination, and human analysts focus on judgment, risk, and outcomes.

github: Dependabot alerts are now assignable to AI agents for remediation

GitHub

2026-04-07

Dependabot alerts can now be assigned to AI coding agents for remediation, including Copilot, Claude, and Codex. This is intended for vulnerabilities that require code changes beyond a simple dependency version update.

March (3)

security: Governing AI agent behavior: Aligning user, developer, role, and organizational intent

Microsoft Security Blog

2026-03-24

The report discusses how to govern AI agent behavior by aligning user, developer, role, and organizational intent for secure enterprise AI adoption.

security: CTI-REALM: A new benchmark for end-to-end detection rule generation with AI agents

Microsoft Security Blog

2026-03-20

Microsoft introduced CTI-REALM, an open-source benchmark for evaluating AI agents on end-to-end detection engineering. It measures whether agents can turn cyber threat intelligence into validated detection rules.

dotnet: Accelerating .NET MAUI Development with AI Agents

.NET

2026-03-19

Microsoft says custom-built AI agents are being used to improve the .NET MAUI contribution workflow, with reported 50–70% reductions in issue resolution time. The post also claims increased test coverage and code quality.