Source: OWASP Top 10 for Agentic Applications (2026)

OWASP's December 2025 list of the ten main risks specific to autonomous agents, from ASI01 Agent Goal Hijack to ASI10 Rogue Agents.

Contents
  • What: community security standard
  • Publisher: OWASP GenAI Security Project, Agentic Security Initiative
  • Date: 2025-12-09; accessed 2026-11-14
  • Original: resource page. Entry names cross-checked against Speakeasy’s explainer because the official PDF was not fetched.

Summary

A companion to the LLM Top 10 for systems that plan, call tools, keep memory and talk to other agents. It is the checklist a red team uses to scope agent-specific tests, and promptfoo exposes it as the owasp:agentic framework filter.

Key claims

  • ASI01 Agent Goal Hijack · ASI02 Tool Misuse and Exploitation · ASI03 Identity and Privilege Abuse · ASI04 Agentic Supply Chain Vulnerabilities · ASI05 Unexpected Code Execution · ASI06 Memory and Context Poisoning · ASI07 Insecure Inter-Agent Communication · ASI08 Cascading Failures · ASI09 Human-Agent Trust Exploitation · ASI10 Rogue Agents.
  • Secondary summaries describe two organising principles, least agency and strong observability, and say the list was built from observed incidents.

Feeds

LLM red teaming

Suggest changes →