Threat Taxonomy

TopAIThreats classifies AI-enabled threats using a multi-dimensional taxonomy. The core hierarchy — domains, patterns, and incidents — is complemented by four stakeholder dimensions (affected groups, exposure pathways, ecosystem positions, and impact level) and six analytical dimensions (causal factors, harm types, assets, attack lifecycle, governance frameworks, and risk levels).

Core hierarchy: Domain → Pattern → Incident

Domains: 8 | Patterns: 48 | Exposure Pathways: 5 | Causal Factors: 15 | Assets: 12 | Harm Types: 7 | Frameworks: 3

Machine-readable: /api/threats.json

DOM-AGT

Agentic & Autonomous Threats

Threats caused by AI systems that act independently, persist over time, or coordinate with other systems.

MIT: Multi-agent risks | EU AI Act: Systemic & autonomy risks (emerging)

PAT-AGT-001 Agent-to-Agent Propagation

PAT-AGT-002 Cascading Hallucinations

PAT-AGT-003 Goal Drift

PAT-AGT-004 Memory Poisoning

PAT-AGT-005 Multi-Agent Coordination Failures

PAT-AGT-007 Specification Gaming

PAT-AGT-006 Tool Misuse & Privilege Escalation

DOM-CTL

Human–AI Control Threats

Threats arising from how humans rely on, defer to, or lose control over AI systems.

MIT: Human-Computer Interaction | EU AI Act: Transparency & oversight requirements

PAT-CTL-001 Deceptive or Manipulative Interfaces

PAT-CTL-002 Implicit Authority Transfer

PAT-CTL-003 Loss of Human Agency

PAT-CTL-004 Overreliance & Automation Bias

PAT-CTL-005 Unsafe Human-in-the-Loop Failures

DOM-ECO

Economic & Labor Threats

Threats that distort markets, labor conditions, or the distribution of economic power.

MIT: Socioeconomic | EU AI Act: Market fairness, systemic risk

PAT-ECO-001 Automation-Induced Job Degradation

PAT-ECO-002 Decision Loop Automation

PAT-ECO-003 Economic Dependency on Black-Box Systems

PAT-ECO-004 Market Manipulation via AI

PAT-ECO-005 Power & Data Concentration

DOM-INF

Information Integrity Threats

Threats that undermine the reliability, authenticity, or shared understanding of information.

MIT: Misinformation | EU AI Act: Manipulation, democratic harm

PAT-INF-006 AI-Enabled Fraud

PAT-INF-001 Consensus Reality Erosion

PAT-INF-002 Deepfake Identity Hijacking

PAT-INF-003 Disinformation Campaigns

PAT-INF-004 Misinformation & Hallucinated Content

PAT-INF-005 Synthetic Media Manipulation

DOM-PRI

Privacy & Surveillance Threats

Threats involving unauthorized inference, tracking, or monitoring of individuals or groups.

MIT: Privacy & Security | EU AI Act: Fundamental rights, GDPR

PAT-PRI-001 Behavioral Profiling Without Consent

PAT-PRI-002 Biometric Exploitation

PAT-PRI-003 Mass Surveillance Amplification

PAT-PRI-004 Re-identification Attacks

PAT-PRI-005 Sensitive Attribute Inference

DOM-SEC

Security & Cyber Threats

AI-enabled attacks that compromise the integrity, confidentiality, or availability of digital systems — through input manipulation, model exploitation, or automated offense.

MIT: Privacy & Security | EU AI Act: Cybersecurity & Robustness

PAT-SEC-001 Adversarial Evasion

PAT-SEC-008 AI Supply Chain Attack

PAT-SEC-002 AI-Morphed Malware

PAT-SEC-009 AI-Powered Social Engineering

PAT-SEC-003 Automated Vulnerability Discovery

PAT-SEC-004 Data Poisoning

PAT-SEC-007 Jailbreak & Guardrail Bypass

PAT-SEC-005 Model Inversion & Data Extraction

PAT-SEC-006 Prompt Injection Attack

DOM-SOC

Discrimination & Social Harm

Threats that result in unfair treatment, exclusion, or social harm to individuals or groups.

MIT: Discrimination & Toxicity | EU AI Act: High-risk systems (employment, credit, education)

PAT-SOC-001 Algorithmic Amplification

PAT-SOC-002 Allocational Harm

PAT-SOC-003 Data Imbalance Bias

PAT-SOC-004 Proxy Discrimination

PAT-SOC-005 Representational Harm

DOM-SYS

Systemic & Catastrophic Risks

Threats that emerge from scale, coupling, and accumulation rather than single failures.

MIT: Long-term / existential | EU AI Act: Systemic risk framing (2026+)

PAT-SYS-001 Accumulative Risk & Trust Erosion

PAT-SYS-002 AI-Assisted Biological Threat Design

PAT-SYS-003 Infrastructure Dependency Collapse

PAT-SYS-004 Lethal Autonomous Weapon Systems (LAWS)

PAT-SYS-005 Strategic Misalignment

PAT-SYS-006 Uncontrolled Recursive Self-Improvement (Hypothetical)

Stakeholder Dimensions

Four dimensions capture who is affected, how they are exposed, who is responsible, and the scale of impact.

TAX-AG

Affected Groups

Who is directly harmed by AI-enabled threats, organized into three categories.

Individuals

Organizations

Systems

Individuals

general-public, workers, children, vulnerable-communities

Organizations

business-organizations, government-institutions, critical-infrastructure-operators, developers-ai-builders

Systems

democratic-institutions, national-security-systems, society-at-large

TAX-EP

Exposure Pathways

How individuals and organizations encounter AI-enabled threats.

Adversarial Targeting Algorithmic Decision Impact Direct Interaction Economic Displacement Infrastructure Dependency

TAX-ECO

Ecosystem Positions

Who caused, enabled, or failed to prevent an AI-enabled threat.

Developers & Providers (developers-providers)

Deployers & Operators (deployers-operators)

Regulators & Public Servants (regulators-public-servants)

Organizational Leaders (organizational-leaders)

Direct Users (direct-users)

Indirectly Affected (indirectly-affected)

TAX-IMP

Impact Level

Scale of harm from an individual incident.

Individual (individual)

Organization (organization)

Sector (sector)

Institution (institution)

Society-Wide (society-wide)

Global (global)

Analytical Taxonomy Dimensions

Additional dimensions classify contributing factors, technologies involved, harm categories, governance frameworks, and risk levels.

TAX-HARM

Harm Types

Categories of harm that AI-enabled threats can inflict on individuals, organizations, and society.

HARM-001 Physical Harm

HARM-002 Financial Harm

HARM-003 Privacy Harm

HARM-004 Discrimination Harm

HARM-005 Reputational Harm

HARM-006 Psychological Harm

HARM-007 Systemic Harm

TAX-FRMW

Governance Frameworks

Regulatory and governance frameworks relevant to AI threat mitigation and compliance.

FRMW-001 NIST AI Risk Management Framework

National Institute of Standards and Technology (NIST)

FRMW-002 EU Artificial Intelligence Act

FRMW-003 ISO/IEC 42001 — AI Management System

International Organization for Standardization / International Electrotechnical Commission

TAX-CF

Causal Factors

Contributing factors that enabled or amplified AI-enabled threats, across four categories.

Malicious Misuse

CAUSE-001 Intentional Fraud CAUSE-002 Adversarial Attack CAUSE-003 Weaponization CAUSE-004 Social Engineering

Design & Development

CAUSE-005 Training Data Bias CAUSE-006 Insufficient Safety Testing CAUSE-007 Hallucination Tendency CAUSE-008 Model Opacity

Deployment & Integration

CAUSE-009 Inadequate Access Controls CAUSE-010 Over-Automation CAUSE-011 Prompt Injection Vulnerability CAUSE-012 Misconfigured Deployment

Systemic & Organizational

CAUSE-013 Regulatory Gap CAUSE-014 Accountability Vacuum CAUSE-015 Competitive Pressure

TAX-ASST

Assets & Technologies

AI technologies and data types involved in documented threat incidents.

Data

ASST-002 Biometric Data ASST-007 Identity Credentials ASST-011 Training Datasets

Model

ASST-006 Foundation Models ASST-009 Large Language Models ASST-010 Recommender Systems ASST-012 Voice Synthesis

System

ASST-001 Autonomous Agents ASST-003 Content Platforms ASST-004 Decision Automation

Infrastructure

ASST-005 Financial Systems ASST-008 Industrial Control Systems

TAX-RISK

Risk Levels

Severity, likelihood, and reversibility assessments applied to incidents and patterns.

Severity

Critical, High, Medium, Low — applied to patterns and incidents

Reversibility

Reversible, Partially Reversible, Irreversible — recoverability of harm