Search results

OWASP LLM Top 10 v2.0 Draft Comment: Evaluation Framework + Defense Findings from Agent Security Sandbox

Status: Open.

#844 In OWASP/www-project-top-10-for-large-language-model-applications;

· TGKDre opened on Jun 17, 2026

Reference Implementation Submission: Agent Trust Bench -- live adversarial payment-flow test corpus (138 profiles, OWASP LLM Top-10 aligned)

Status: Open.

#839 In OWASP/www-project-top-10-for-large-language-model-applications;

· chopmob-cloud opened on May 30, 2026

🚨 [BUG] - Pages 39 to 41 are completely malformed in the PDF assets (OWASP-Top-10-for-LLMs-v2025.pdf)

bugFactual inaccuracies, broken links, typos, etcFactual inaccuracies, broken links, typos, etc

Status: Open.

#834 In OWASP/www-project-top-10-for-large-language-model-applications;

· Vincent-Stragier opened on May 22, 2026

💡 [REQUEST] - [LLM01] Add Bulwark — open-source defense framework with prompt-injection detection, RBAC, and audit

Status: Open.

#830 In OWASP/www-project-top-10-for-large-language-model-applications;

· anilatambharii opened on May 3, 2026

Proposal: OWASP Top 10 for AI-Generated Code (AGC)

Status: Open.

#826 In OWASP/www-project-top-10-for-large-language-model-applications;

· razashariff opened on Apr 17, 2026

Reference Implementation Submission: AASTF — Agentic AI Security Testing Framework

Status: Open.

#825 In OWASP/www-project-top-10-for-large-language-model-applications;

· anonymousAAK opened on Apr 15, 2026

Proposal: Boundary-to-Boundary Invariant Survival for Agentic Execution Governance

Status: Open.

#817 In OWASP/www-project-top-10-for-large-language-model-applications;

· QueBallSharken opened on Apr 6, 2026

Proposal: LLM11 - Observation Framing Attack (OFA)

Status: Open.

#813 In OWASP/www-project-top-10-for-large-language-model-applications;

· NoDataFound opened on Mar 30, 2026

Agent Passport System: cryptographic enforcement layer for ASI-03, ASI-07, ASI-08, ASI-10

Status: Open.

#812 In OWASP/www-project-top-10-for-large-language-model-applications;

· aeoess opened on Mar 27, 2026

Case Study: Agent Memory Chain Attack - Persistent Code Execution via Prompt Injection in Autonomous Agents

Status: Open.

#811 In OWASP/www-project-top-10-for-large-language-model-applications;

· Preventnetworkhacking opened on Mar 26, 2026

Proposed New Category: Postural Manipulation — Semantically Benign Behavioral Orientation Attack

Status: Open.

#807 In OWASP/www-project-top-10-for-large-language-model-applications;

· adavidson510 opened on Mar 18, 2026

MCP attack mechanics missing from LLM01 and LLM06: five lab-confirmed patterns not currently documented

Status: Open.

#806 In OWASP/www-project-top-10-for-large-language-model-applications;

· aminrj opened on Mar 17, 2026