OWASP Top 10 for LLM Applications 2025
OWASP Top 10 security risks specific to Large Language Model (LLM) applications. Identifies the most critical vulnerabilities in AI/LLM systems including prompt injection, data poisoning, and excessive agency. Published by the OWASP GenAI Security Project.
Framework summaries on this platform are AI-assisted interpretations for educational and compliance planning purposes. They do not reproduce or replace the official standards. Refer to the authoritative source for the definitive text. Framework names and trademarks belong to their respective organisations.
OWASP content is used under the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0). Original material © OWASP Foundation. See owasp.org for the authoritative source.
Framework Domains (4)
Data and Training Security
Risks related to training data integrity and information disclosure
| Code | Title |
|---|---|
| LLM02-2025 | Sensitive Information Disclosure |
| LLM04-2025 | Data Poisoning |
| LLM08-2025 | Vector and Embedding Weaknesses |
Input and Prompt Security
Risks related to how inputs are processed by LLM systems
| Code | Title |
|---|---|
| LLM01-2025 | Prompt Injection |
| LLM07-2025 | System Prompt Leakage |
Output and Integration Security
Risks related to LLM outputs and downstream system interactions
| Code | Title |
|---|---|
| LLM05-2025 | Improper Output Handling |
| LLM06-2025 | Excessive Agency |
| LLM09-2025 | Misinformation |
Supply Chain and Resource Security
Risks related to third-party components and resource consumption
| Code | Title |
|---|---|
| LLM03-2025 | Supply Chain Vulnerabilities |
| LLM10-2025 | Unbounded Consumption |
Maps to 532 other frameworks
Frequently Asked Questions
What is OWASP Top 10 for LLM Applications 2025?
OWASP Top 10 for LLM Applications 2025 is a compliance framework from International with 4 domains and 10 controls. OWASP Top 10 security risks specific to Large Language Model (LLM) applications. Identifies the most critical vulnerabilities in AI/LLM systems including prompt injection, data poisoning, and excessive agency. Published by the OWASP GenAI Security Project. It is used by organisations to establish and maintain compliance with industry standards and regulatory requirements.
How many controls does OWASP Top 10 for LLM Applications 2025 have?
OWASP Top 10 for LLM Applications 2025 has 10 controls organised across 4 domains. The largest domains are Data and Training Security (3 controls), Output and Integration Security (3 controls), Input and Prompt Security (2 controls). Each control defines specific requirements that organisations must implement to achieve compliance.
What frameworks does OWASP Top 10 for LLM Applications 2025 map to?
OWASP Top 10 for LLM Applications 2025 maps to 532 other compliance frameworks. The top mapping partners are WCO Authorised Economic Operator (AEO) Framework (50% coverage), ASD Information Security Manual (ISM) (50% coverage), NIS2 Directive Implementing Acts (50% coverage). Use our comparison tool to explore control-level mappings between frameworks.
How do I get started with OWASP Top 10 for LLM Applications 2025 compliance?
Start your OWASP Top 10 for LLM Applications 2025 compliance journey by running a self-assessment on our platform to identify your current compliance posture. Our AI advisory can answer specific questions about OWASP Top 10 for LLM Applications 2025 requirements, and cross-framework mapping helps you leverage existing controls from other frameworks you may already comply with. Create a free account to access all 10 controls and track your progress.
Start Your Compliance Journey
Create a free account to run self-assessments, get AI advisory, and track your compliance progress across 692 frameworks.
Get Started Free →Free forever — no credit card required