Updating the taxonomy of failure modes in agentic AI systems: What a year of red teaming taught us
A surge in real-world attacks against agentic AI systems is reshaping how we think about risk.
Actionable threat insights
Get timely insights into emerging vulnerabilities, first‑finder discoveries, and evolving cyberattacker behaviors. Explore deep research and real‑world Microsoft Defender scenarios that show how proactive detection and quick action help organizations prevent compromise.
Refine results
Topic
Threat intelligence
Products and services
Publish date
-
-
Preinstall to persistence: Inside the Red Hat npm Miasma credential-stealing campaign
A large-scale npm supply chain attack compromised over 90 versions of @redhat-cloud-services packages, silently infecting CI/CD environments and developer systems. -
Malicious npm packages abuse dependency confusion to profile developer environments
A dependency confusion campaign leveraged 33 malicious npm packages to collect reconnaissance data from developer and build environments. -
Typosquatted npm packages used to steal cloud and CI/CD secrets
The Mini Shai-Hulud campaign used malicious npm packages to target cloud and CI/CD credentials across developer environments. Retain Microsoft Security Experts
Microsoft Security Experts are now available to strengthen your team with managed security services. Learn how to defend against threats with security experts.
-
From poisoned search results to GPU mining: A cryptojacking campaign abusing ScreenConnect and Microsoft .NET utilities
Microsoft exposes a cryptojacking campaign using SEO poisoning and ScreenConnect to target high-performance PCs, with malicious sites also surfaced through AI chatbots. -
From edge appliance to enterprise compromise: Multi-stage Linux intrusion via F5 and Confluence
A multi-stage attack on Linux devices began with an exposed F5 BIG-IP edge appliance and pivoted to an internal Confluence server for credential theft and identity compromise. -
Mini Shai Hulud: Compromised @antv npm packages enable CI/CD credential theft
Compromised @antv npm packages deploy the Mini Shai-Hulud payload to steal CI/CD secrets from Linux-based automation environments. -
Introducing RAMPART and Clarity: Open source tools to bring safety into Agent development workflow
The AI systems shipping inside enterprises today are fundamentally different from the ones we were building even two years ago, because they have moved well past answering questions and into accessing your email, retrieving records from your CRM, writing and executing code, and taking actions on your behalf across dozens of connected systems. Modernize your Security Operations Center with Microsoft Sentinel
Microsoft Sentinel is a cloud-native SIEM solution powered by AI and automation that delivers intelligent security analytics across your entire enterprise.
-
How Storm-2949 turned a compromised identity into a cloud-wide breach
Storm-2949 turned stolen credentials into a cloud-wide breach, moving from identity compromise to large-scale data theft without using malware. -
Defense in depth for autonomous AI agents
As AI agents gain autonomy, defense in depth must evolve, with application-layer design, identity, and human oversight at the center. -
When configuration becomes a vulnerability: Exploitable misconfigurations in AI apps
Exposed UIs, weak authentication, and risky defaults could turn cloud-native AI apps on Kubernetes into potential targets by threat actors. -
Accelerating detection engineering using AI-assisted synthetic attack logs generation
What if you could generate realistic attack telemetry on demand? Explore research methods that translate attacker behaviors (TTPs) into synthetic logs that can trigger detections at scale and without sensitive data.
