Appearance
Use OWASP Top 10 for Agentic Applications in Modulos
Modulos has one Framework for the OWASP Top 10 for Agentic Applications, Version 2026: MFF-14 OWASP - Top 10 for Agentic Applications. You add MFF-14 to an AI System. Modulos has no Framework for the Agentic Top 10 at the Organization level. See Use in Modulos.
The OWASP Top 10 for Agentic Applications is security information from OWASP. It is not law. See OWASP Top 10 for Agentic Applications.
The sections below tell you how to add MFF-14, find each Requirement's OWASP section, and do the Requirement Owner's Review.
Add the Framework
For the OWASP Top 10 for Agentic Applications, MFF-14 gives one Requirement for each OWASP risk. You do not make these Requirements.
These conditions apply when you add MFF-14:
- Only a person with the permissions to add Frameworks to the AI System can add
MFF-14. - If Modulos has a new version of a Framework that the AI System uses, you cannot add a Framework. First, change the AI System to the new version of that Framework.
- While Modulos does a refresh or a freeze of a Framework of the AI System, you cannot add a Framework.
For these conditions, see AI System Settings.
- If your agentic application has no AI System, make one. See Register an AI System.
- Open the AI System of your agentic application.
- Open Settings.
- Open Frameworks.
- Add
MFF-14.
Requirement mapping
In MFF-14, each Requirement refers to one section of the OWASP Top 10 for Agentic Applications, Version 2026.
| Requirement | Requirement name | OWASP entry | Section for the risk |
|---|---|---|---|
MRF-245 | ASI01:2026 - Agent Goal Hijack | ASI01 | Agent Goal Hijack |
MRF-246 | ASI02:2026 - Tool Misuse and Exploitation | ASI02 | Tool Misuse and Exploitation |
MRF-247 | ASI03:2026 - Identity and Privilege Abuse | ASI03 | Identity and Privilege Abuse |
MRF-248 | ASI04:2026 - Agentic Supply Chain Vulnerabilities | ASI04 | Agentic Supply Chain Vulnerabilities |
MRF-249 | ASI05:2026 - Unexpected Code Execution (RCE) | ASI05 | Unexpected Code Execution (RCE) |
MRF-250 | ASI06:2026 - Memory & Context Poisoning | ASI06 | Memory & Context Poisoning |
MRF-251 | ASI07:2026 - Insecure Inter-Agent Communication | ASI07 | Insecure Inter-Agent Communication |
MRF-252 | ASI08:2026 - Cascading Failures | ASI08 | Cascading Failures |
MRF-253 | ASI09:2026 - Human-Agent Trust Exploitation | ASI09 | Human-Agent Trust Exploitation |
MRF-254 | ASI10:2026 - Rogue Agents | ASI10 | Rogue Agents |
The mapping does not show that MFF-14 has measures for each risk of your agentic application. It also does not show that MFF-14 has all measures that OWASP recommends. See Control overlap.
Review the Requirements
For the OWASP Top 10 for Agentic Applications in MFF-14, the Requirement Owner does the Review of each Requirement. Requirement fulfillment is not automatic.
As the Requirement Owner, do these steps for each Requirement:
- Make sure that all Controls of the Requirement have a final status.
- Examine the completed Controls of the Requirement.
- If your Review shows fulfillment, record fulfillment in Modulos.
For the conditions of the Review, see Requirements.
Sources
- OWASP GenAI Security Project, OWASP Top 10 for Agentic Applications for 2026, Version 2026, December 2025: the entries ASI01 to ASI10.
- For Modulos, see AI System Settings, Register an AI System, and Requirements.
Disclaimer
This page gives general information. It is not legal advice.