CalvinBackup/www-project-ai-testing-guide
1
0
Fork 0
mirror of https://github.com/OWASP/www-project-ai-testing-guide.git synced 2026-08-30 04:40:35 +02:00
Code Issues Packages Projects Releases Wiki Activity

Update AITG-MOD-02_Testing_for_Runtime_Model_Poisoning.md

Browse Source
This commit is contained in:
Matteo Meucci
2025-11-23 13:49:44 +01:00
committed by GitHub
parent c4fe008037
commit a5485eab40
1 changed files with 3 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Document/content/tests/AITG-MOD-02_Testing_for_Runtime_Model_Poisoning.md
+3 -6
View File
@@ -29,13 +29,10 @@ This test identifies vulnerabilities associated with runtime model poisoning, wh
- **Weight Feedback Based on Trust**: Implement a trust score for users. Feedback from new or low-trust users should have a much smaller impact on model updates than feedback from long-standing, high-trust users.
- **Periodically Retrain from Scratch**: To wash out any poison that may have accumulated, periodically discard the online model and retrain a new one from scratch using a clean, verified, and comprehensive dataset.
### Suggested Tools for this Specific Test
- **Adversarial Robustness Toolbox (ART)**
- Provides capabilities for crafting and defending against runtime poisoning attacks, particularly for deep learning models.
- Tool Link: [ART on GitHub](https://github.com/Trusted-AI/adversarial-robustness-toolbox)
### Suggested Tools
- **Adversarial Robustness Toolbox (ART)**: Provides capabilities for crafting and defending against runtime poisoning attacks, particularly for deep learning models - [ART on GitHub](https://github.com/Trusted-AI/adversarial-robustness-toolbox)
- **Custom Scripts with Scikit-learn**: As demonstrated above, `scikit-learn`'s `partial_fit` method is excellent for simulating online learning and testing runtime poisoning concepts.
- **River**: A Python library specifically designed for online machine learning, providing a more advanced environment for simulating these attacks.
- Tool Link: [River on GitHub](https://github.com/online-ml/river)
- **River**: A Python library specifically designed for online machine learning, providing a more advanced environment for simulating these attacks - [River on GitHub](https://github.com/online-ml/river)
### References
- OWASP Top 10 for LLM Applications 2025. "LLM04: Data and Model Poisoning." OWASP, 2025. [Link](https://genai.owasp.org/llmrisk/llm042025-data-and-model-poisoning/)