The Future of AI Agent Security Is Guardrails
AI agents introduce new security risks like prompt injection and data exfiltration. Learn how guardrails, hook-based controls, and Arcade’s Contextual Access secure AI agent tool calls in real time.

4276 articles
4276 ARTICLES
AI agents introduce new security risks like prompt injection and data exfiltration. Learn how guardrails, hook-based controls, and Arcade’s Contextual Access secure AI agent tool calls in real time.


Linky’s Matchmaker converts your Dockerfile to Chainguard Containers, recommending secure, minimal base images and generating an updated file in minutes

Fulfillment Dashboard gives Chainguard customers real-time visibility, tracking, and community voting for new secure container image requests.

Are “Skill Scanners” on ClawHub actually safe? We tested popular community tools like Skill Defender and SkillGuard against real malware. The results were alarming.

The latest tag isn’t unsafe by default — pin images to digests for reproducible, secure updates while staying current with automated workflows.

Read Microsoft’s new Cyber Pulse report for straightforward, practical insights and guidance on new cybersecurity risks. The post 80% of Fortune 500 use active AI Agents: Observability, governance, and security shape the new frontier appeared first on Microsoft Security Blog.

Breaking: Snyk researchers uncover a malicious “Google” skill on ClawHub that tricks users into installing malware via a fake OpenClaw dependency. Learn how the attack works and how to protect your AI agents.

Bring Wiz Issues directly into Backstage, so developers can act on security issues in the tools they use everyday

Discover how 7.1% of AI agent skills are designed to leak secrets, PII, and API keys through LLM context. Learn to defend with Evo & mcp-scan.

Welcome to the Top 10 Web Hacking Techniques of 2025, the 19th edition of our annual community-powered effort to identify the most innovative must-read web security research published in the last year

Snyk’s ToxicSkills research reveals 36% of AI agent skills contain security flaws, including 1,467 vulnerable skills and active malicious payloads targeting OpenClaw, Claude Code, and Cursor users.

Chainguard can help your engineering team accelerate sales velocity by reducing security friction and enabling engineering to be a growth accelerator.
Cookies
We use analytics cookies (Google Analytics) to improve this site. Accept to allow them. Privacy Policy