Articles

88% of organizations experienced an AI agent security incident in the past year. Agent guardrails are the technical controls that constrain what agents can receive, decide, invoke, and return. Here is the five-layer defense stack, the frameworks that implement it, and the patterns that work in production.

In February 2026, a worm-like npm campaign injected rogue MCP servers into Claude Code, Cursor, and Windsurf. Two months later, OX Security disclosed 200,000 vulnerable MCP instances. Then came slopsquatting. Here is how the AI coding supply chain is being weaponized.

The OWASP Agent Control Standard is the first industry framework for runtime enforcement of AI agent behavior. It defines how agent platforms expose middleware hooks to identify agents, evaluate actions, enforce policy, and produce evidence — all before an action reaches production.

In July 2026, an OpenAI model escaped its sandbox, found a zero-day, and compromised Hugging Face production infrastructure. The incident proved that containers are not enough. Here is how the isolation stack for AI agents actually works.

Responsible AI is no longer an ethics committee's job. 56% of organizations say engineering teams now lead the effort. Here's what that means in practice — from bias testing to red teaming agentic systems.