All Records

Compromising privacy by leaking sensitive information

Compromising privacy by leaking sensitive information is an AI risk in 2. Privacy & Security focused on 2.1 > Compromise of privacy by leaking or correctly i...

PoisonGPT - AI Case Study

PoisonGPT - AI Case Study is an AI security case study covering type: exercise; target: HuggingFace Users; actor: Mithril Security Researchers. It includes 7...

1.1 > Unfair discrimination and misrepresentation AI Risks

1.1 > Unfair discrimination and misrepresentation groups 77 AI risk records under 1. Discrimination & Toxicity.

Pursuing Consistent Context

Pursuing Consistent Context is an AI risk in 3. Misinformation focused on 3.1 > False or misleading information. It is most relevant during 2 - Post-deployment.

Supply Chain Compromise via Poisoned ClawdBot Skill - AI Case Study

Supply Chain Compromise via Poisoned ClawdBot Skill - AI Case Study is an AI security case study covering type: exercise; target: ClawdBot (now OpenClaw); ac...

Data Exfiltration via an MCP Server used by Cursor - AI Case Study

Data Exfiltration via an MCP Server used by Cursor - AI Case Study is an AI security case study covering type: exercise; target: Cursor; actor: Backslash Sec...

Microsoft Azure Service Disruption - AI Case Study

Microsoft Azure Service Disruption - AI Case Study is an AI security case study covering type: exercise; target: Internal Microsoft Azure Service; actor: Mic...

Privacy Violation

Privacy Violation is an AI risk in 2. Privacy & Security focused on 2.1 > Compromise of privacy by leaking or correctly inferring sensitive information. It i...

Cybersecurity

Cybersecurity is an AI risk in 2. Privacy & Security focused on 2.0 > Privacy & Security. It is most relevant during 3 - Other.

Privacy Harms

Privacy Harms is an AI risk in 2. Privacy & Security focused on 2.1 > Compromise of privacy by leaking or correctly inferring sensitive information. It is mo...

Embed Malware - AI Security Technique

Embed Malware is AI attack method AML.T0018.002 with evidence level: realized. It includes 1 mitigation; 2 case studies; 5 related AI risks.

AI Model Tampering via Supply Chain Attack - AI Case Study

AI Model Tampering via Supply Chain Attack - AI Case Study is an AI security case study covering type: exercise; target: Private Container Registries; actor...

Extract LLM System Prompt - AI Security Technique

Extract LLM System Prompt is AI attack method AML.T0056 with evidence level: feasible. It includes 3 mitigations; 12 related AI risks.

Exposed ClawdBot Control Interfaces Leads to Credential Access and Execution - AI Case Study

Exposed ClawdBot Control Interfaces Leads to Credential Access and Execution - AI Case Study is an AI security case study covering type: exercise; target: Cl...

Reputational Harm - AI Security Technique

Reputational Harm is AI attack method AML.T0048.001 with evidence level: demonstrated. It includes 1 case study; 9 related AI risks.

Restrict Number of AI Model Queries - AI Mitigation

Restrict Number of AI Model Queries is an AI mitigation for Business and Data Understanding, Deployment, Monitoring and Maintenance, grouped as Technical - C...

Data Protection/Privacy

Data Protection/Privacy is an AI risk in 2. Privacy & Security focused on 2.1 > Compromise of privacy by leaking or correctly inferring sensitive information...

Face Identification System Evasion via Physical Countermeasures - AI Case Study

Face Identification System Evasion via Physical Countermeasures - AI Case Study is an AI security case study covering type: exercise; target: Commercial Face...

AI Agent Context Poisoning - AI Security Technique

AI Agent Context Poisoning is AI attack method AML.T0080 with evidence level: demonstrated. It includes 1 mitigation; 13 related AI risks.

Datasets - AI Security Technique

Datasets is AI attack method AML.T0002.000 with evidence level: demonstrated. It includes 1 mitigation; 6 case studies.