PPhying中文
Curated security research, vulnerabilities, advisories and tools for practitioners.

Advancing AI Red Teaming: Agentic Taxonomies and Governance-Ready Reporting

hackerone.com | blog | #ai-security | #agentic-ai | #compliance | #prompt-injection | #owasp | #ai-red-teaming | #eu-ai-act | #risk-governance
📄 Raw .md⬇️ Download .md🔗 Original ↗

Summary

AI Red Teaming now maps validated agentic exploit paths to OWASP, NIST, ISO, and the EU AI Act. Get governance-ready reporting without spreadsheets.
Published
Collected

Related coverage

blog · bugcrowd.com
What’s changed in AI security over the past two years
Bugcrowd compares its 2024 and 2026 AI security guides: 80% of Fortune 500 run autonomous agents, making memory poisoning and plan hijacking real attack classes — plus regulation and 'AI slop'.
blog · crowdstrike.com
Agents of Chaos: A New $100K Agentic Security Challenge
CrowdStrike launched Agents of Chaos, a $100,000 online game and AI red teaming competition in which players manipulate real AI agents using prompt injection and other adversarial techniques.
blog · blog.sentry.security
Semantic vs. Token-Based LLM Injections
Contrasting two LLM injection classes: semantic attacks use natural language to override instructions; token-based attacks exploit tokenizer quirks like reserved delimiters and adversarial suffixes.
blog · blog.sentry.security
LLM API Misconfigurations
How LLM API misconfigurations break trust boundaries: exposed messages[], privileged roles, model choice, or generation parameters enable role injection, parameter tampering, and runaway token usage.
blog · blog.sentry.security
ChatML Role Injection (CRI)
Attackers can inject system roles into ChatML message arrays to override developer instructions, invoke hidden tools, and pad payloads for denial-of-wallet DoS; strict schema validation is key.
blog · blog.sentry.security
AI Red Team: Safety vs. Security
AI red teaming should separate security from safety: security covers compromise of AI systems like prompt injection, while safety addresses harmful misuse such as LLM-enabled malware.
Back