# 3. AI Testing Guide Framework Based on the Threat modeling performed at Chapter 2, we can now define a structured framework that maps the AI Architecture threats to concrete test cases. This project aims to bridge traditional cybersecurity, MLOps testing, and Responsible AI assessments under a unified structure. Each test case is categorized under one of four pillars: - 🟦 [**AI Application Testing**](3.1_AI_Application_Testing.md) - 🟪 [**AI Model Testing**](3.2_AI_Model_Testing.md) - 🟩 [**AI Infrastructure Testing**](3.3_AI_Infrastructure_Testing.md) - 🟨 [**AI Data Testing**](3.4_AI_Data_Testing.md) Before starting the analysis, it is important to take into account the limitations of this type of testing and consider the possibility of moving from a black-box approach to a grey-box or white-box approach, which requires additional information. **Limitations and requirements** are described in the chapter **[Testing Limitations and Requirements](3.0_Testing_Limitations_and_Requirements.md)**. ## 🟦 AI Application Testing | Test ID | Test Name & Link | Threat Source | Domain(s) | |-------------|----------------------------------------------------------------------------------------|-------------------------|---------------------------------------| | AITG-APP-01 | [Testing for Prompt Injection](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-01_Testing_for_Prompt_Injection.md) | OWASP Top 10 LLM 2025 | Security | | AITG-APP-02 | [Testing for Indirect Prompt Injection](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-02_Testing_for_Indirect_Prompt_Injection.md)| OWASP Top 10 LLM 2025 | Security | | AITG-APP-03 | [Testing for Sensitive Data Leak](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-03_Testing_for_Sensitive_Data_Leak.md) | OWASP Top 10 LLM 2025 | Security, Privacy | | AITG-APP-04 | [Testing for Input Leakage](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-04_Testing_for_Input_Leakage.md) | OWASP Top 10 LLM 2025 | Security, Privacy | | AITG-APP-05 | [Testing for Unsafe Outputs](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-05_Testing_for_Unsafe_Outputs.md) | OWASP Top 10 LLM 2025 | Security, RAI | | AITG-APP-06 | [Testing for Agentic Behavior Limits](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-06_Testing_for_Agentic_Behavior_Limits.md) | OWASP Top 10 LLM 2025 | Security, Trustworthy AI | | AITG-APP-07 | [Testing for Prompt Disclosure](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-07_Testing_for_Prompt_Disclosure.md) | OWASP Top 10 LLM 2025 | Security, Privacy | | AITG-APP-08 | [Testing for Embedding Manipulation](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-08_Testing_for_Embedding_Manipulation.md) | OWASP Top 10 LLM 2025 | Security | | AITG-APP-09 | [Testing for Model Extraction](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-09_Testing_for_Model_Extraction.md) | OWASP AI Exchange | Security | | AITG-APP-10 | [Testing for Harmful Content Bias](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-10_Testing_for_Harmful_Content_Bias.md) | OWASP Top 10 LLM 2025 | RAI | | AITG-APP-11 | [Testing for Hallucinations](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-11_Testing_for_Hallucinations.md) | Trustworthy AI | Trustworthy AI | | AITG-APP-12 | [Testing for Toxic Output](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-12_Testing_for_Toxic_Output.md) | Responsible AI | RAI | | AITG-APP-13 | [Testing for Over-Reliance on AI](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-13_Testing_for_Over-Reliance_on_AI.md) | Responsible AI | RAI, Trustworthy AI | | AITG-APP-14 | [Testing for Explainability and Interpretability](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-APP-14_Testing_for_Explainability_and_Interpretability.md) | Responsible AI | RAI, Trustworthy AI | --- ## 🟪 AI Model Testing | Test ID | Test Name & Link | Threat Source | Domain(s) | |-------------|-----------------------|-------------------------|-----------| | AITG-MOD-01 | [Testing for Evasion Attacks](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-MOD-01_Testing_for_Evasion_Attacks.md) | OWASP AI Exchange | Security | | AITG-MOD-02 | [Testing for Runtime Model Poisoning](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-MOD-02_Testing_for_Runtime_Model_Poisoning.md) | OWASP Top 10 LLM 2025 | Security | | AITG-MOD-03 | [Testing for Poisoned Training Sets](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-MOD-03_Testing_for_Poisoned_Training_Sets.md) | OWASP Top 10 LLM 2025 | Security | | AITG-MOD-04 | [Testing for Membership Inference](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-MOD-04_Testing_for_Membership_Inference.md) | OWASP AI Exchange | Privacy | | AITG-MOD-05 | [Testing for Inversion Attacks](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-MOD-05_Testing_for_Inversion_Attacks.md) | OWASP AI Exchange | Privacy | | AITG-MOD-06 | [Testing for Robustness to New Data](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-MOD-06_Testing_for_Robustness_to_New_Data.md) | Trustworthy AI | Trustworthy AI | | AITG-MOD-07 | [Testing for Goal Alignment](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-MOD-07_Testing_for_Goal_Alignment.md) | Trustworthy AI | Trustworthy AI | ## 🟩 AI Infrastructure Testing | Test ID | Test Name & Link | Threat Source | Domain(s) | |-------------|-----------------------|-------------------------|-----------| | AITG-INF-01 | [Testing for Supply Chain Tampering](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-INF-01_Testing_for_Supply_Chain_Tampering.md) | OWASP Top 10 LLM 2025 | Security | | AITG-INF-02 | [Testing for Resource Exhaustion](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-INF-02_Testing_for_Resource_Exhaustion.md) | OWASP Top 10 LLM 2025 | Security | | AITG-INF-03 | [Testing for Plugin Boundary Violations](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-INF-03_Testing_for_Plugin_Boundary_Violations.md) | Trustworthy AI | Trustworthy AI | | AITG-INF-04 | [Testing for Capability Misuse](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-INF-04_Testing_for_Capability_Misuse.md) | Responsible AI | RAI, Trustworthy AI | | AITG-INF-05 | [Testing for Fine-tuning Poisoning](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-INF-05_Testing_for_Fine-tuning_Poisoning.md) | OWASP Top 10 LLM 2025 | Security | | AITG-INF-06 | [Testing for Dev-Time Model Theft](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-INF-06_Testing_for_Dev-Time_Model_Theft.md) | OWASP AI Exchange | Security, Privacy | --- ## 🟨 AI Data Testing | Test ID | Test Name & Link | Threat Source | Domain(s) | |-------------|-----------------------|-------------------------|-----------| | AITG-DAT-01 | [Testing for Training Data Exposure](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-DAT-01_Testing_for_Training_Data_Exposure.md) | OWASP AI Exchange | Privacy | | AITG-DAT-02 | [Testing for Runtime Exfiltration](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-DAT-02_Testing_for_Runtime_Exfiltration.md) | OWASP AI Exchange | Security, Privacy | | AITG-DAT-03 | [Testing for Dataset Diversity & Coverage](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-DAT-03_Testing_for_Dataset_Diversity_and_Coverage.md) | Responsible AI | RAI | | AITG-DAT-04 | [Testing for Harmful Content in Data](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-DAT-04_Testing_for_Harmful_Content_in_Data.md) | Responsible AI | RAI | | AITG-DAT-05 | [Testing for Data Minimization & Consent](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/tests/AITG-DAT-05_Testing_for_Data_Minimization_and_Consent.md) | Trustworthy AI | Privacy, Trustworthy AI | NEXT: 3.0 [Testing Limitations and Requirements](https://github.com/OWASP/www-project-ai-testing-guide/blob/main/Document/content/3.0_Testing_Limitations_and_Requirements.md) [Table of Content](README.md)