Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Initially published by the Open Worldwide Application Security Project (OWASP) in 2023, the Top 10 for LLM Application Security list seeks to bridge the gap between traditional application security and the unique threats related to large language models (LLMs). Even where the vulnerabilities listed have the same names, the Top 10 for LLM Application Security focuses on how threat actors can exploit LLMs in new ways and potential remediation strategies that developers can implement.

Security teams deploying AI agents into Kubernetes know they need behavioral baselines. The concept is straightforward: define what “normal” looks like for each agent, then detect when behavior drifts in ways that suggest compromise. The problem is that AI agents are designed to change. A model update alters inference latency. A prompt revision shifts tool-calling sequences. A new MCP integration adds API destinations nobody flagged during the last security review.

Everyone is calling Claude Mythos a watershed moment. I’d like to offer a slightly different take. Not because the capability isn’t real, it is. But if Mythos is the moment that finally convinced your organization that rapid vulnerability discovery is an existential threat, you’ve been watching the wrong thing. We saw this coming. Vulnerability Management has been moving in this direction for years, and we built Nucleus with this trajectory in mind. What surprises me is the surprise.

AI is changing how attacks happen, and how fast they happen. Seemplicity’s 2026 State of Exposure Management report shows why most security teams aren’t struggling to find risk, but to fix it quickly enough. Based on insights from 300 security leaders, it highlights where remediation breaks down, how AI is being used today, and why execution is becoming the real bottleneck.

On April 6, 2026, Microsoft Defender Security Research published an advisory detailing a large-scale phishing campaign that leverages the OAuth Device Code Authentication flow to compromise Microsoft 365 accounts across organizations globally. This campaign represents a significant evolution from manual social engineering to fully automated, AI-driven attack infrastructure.

An AI just found critical vulnerabilities that survived decades of human review. If your security program isn’t built for this moment, it’s already behind. Surprise! An AI just did what your security team couldn’t.