All Records

AI Agent Tools - AI Security Technique

AI Agent Tools is AI attack method AML.T0085.001 with evidence level: demonstrated. It includes 5 mitigations; 3 case studies.

AI Service Proxies - AI Security Technique

AI Service Proxies is AI attack method AML.T0008.005 with evidence level: realized. It includes 1 case study; 5 related AI risks.

Create Proxy AI Model - AI Security Technique

Create Proxy AI Model is AI attack method AML.T0005 with evidence level: demonstrated. It includes 5 mitigations; 3 case studies.

Control Access to AI Models and Data at Rest - AI Mitigation

Control Access to AI Models and Data at Rest is an AI mitigation for Business and Data Understanding, Data Preparation, ML Model Evaluation, ML Model Enginee...

AI ClickFix: Hijacking Computer-Use Agents Using ClickFix - AI Case Study

AI ClickFix: Hijacking Computer-Use Agents Using ClickFix - AI Case Study is an AI security case study covering type: exercise; target: Claude Computer-Use A...

Arbitrary Code Execution with Google Colab - AI Case Study

Arbitrary Code Execution with Google Colab - AI Case Study is an AI security case study covering type: exercise; target: Google Colab; actor: Tony Piazza. It...

4.3 > Fraud, scams, and targeted manipulation AI Risks

4.3 > Fraud, scams, and targeted manipulation groups 66 AI risk records under 4. Malicious Actors & Misuse.

Anthropomorphising systems can lead to overreliance or unsafe use

Anthropomorphising systems can lead to overreliance or unsafe use is an AI risk in 5. Human-Computer Interaction focused on 5.1 > Overreliance and unsafe use...

Models - AI Security Technique

Models is AI attack method AML.T0002.001 with evidence level: demonstrated. It includes 2 mitigations; 4 case studies.

Discrimination, toxicity, and bias

Discrimination, toxicity, and bias is an AI risk in 1. Discrimination & Toxicity focused on 1.1 > Unfair discrimination and misrepresentation. It is most rel...

Morris II Worm: RAG-Based Attack - AI Case Study

Morris II Worm: RAG-Based Attack - AI Case Study is an AI security case study covering type: exercise; target: RAG-based e-mail assistant; actor: Stav Cohen...

Data Destruction via AI Agent Tool Invocation - AI Security Technique

Data Destruction via AI Agent Tool Invocation is AI attack method AML.T0101 with evidence level: realized. It includes 6 mitigations; 2 case studies.

Living Off AI: Prompt Injection via Jira Service Management - AI Case Study

Living Off AI: Prompt Injection via Jira Service Management - AI Case Study is an AI security case study covering type: exercise; target: Atlassian MCP, Jira...

Prompt leaking

Prompt leaking is an AI risk in 2. Privacy & Security focused on 2.2 > AI system security vulnerabilities and attacks. It is most relevant during 3 - Other.

Exfiltration via AI Inference API - AI Security Technique

Exfiltration via AI Inference API is AI attack method AML.T0024 with evidence level: realized. It includes 3 mitigations; 6 related AI risks.

Data Exfiltration via Remote Poisoned MCP Tool - AI Case Study

Data Exfiltration via Remote Poisoned MCP Tool - AI Case Study is an AI security case study covering type: exercise; target: Model Context Protocol; actor: I...

Cyberspace risks (Risks of information leakage due to improper usage)

Cyberspace risks (Risks of information leakage due to improper usage) is an AI risk in 2. Privacy & Security focused on 2.1 > Compromise of privacy by leakin...

Risks from AI systems (Risks of exploitation through defects and backdoors)

Risks from AI systems (Risks of exploitation through defects and backdoors) is an AI risk in 2. Privacy & Security focused on 2.2 > AI system security vulner...

Masquerading - AI Security Technique

Masquerading is AI attack method AML.T0074 with evidence level: realized. It includes 4 case studies.

External Harms - AI Security Technique

External Harms is AI attack method AML.T0048 with evidence level: realized. It includes 4 case studies.