Public Database
Case Studies
Every approved AI failure case, classified against the AI Blindspot Framework. New to AIBlindspot? Start with the overview or the methodology.
Chaotic Dynamics in Multi-Agent AI Systems Grow Harder to Predict at Scale
Recent case. Full summary visible to registered users — sign in to read.
Multi-Agent AI Systems Produce Unstable Cyclic Behaviour Under Standard Learning Rules
Recent case. Full summary visible to registered users — sign in to read.
Abrupt Behavioural Phase Transitions in Multi-Agent AI Systems
Recent case. Full summary visible to registered users — sign in to read.
AI Agents Unable to Form Credible Commitments in Collaborative Tasks
Recent case. Full summary visible to registered users — sign in to read.
Emergent Agency Risk in Composed Multi-Agent AI Systems
Recent case. Full summary visible to registered users — sign in to read.
Multi-Agent AI Systems Combining Capabilities to Defeat Security Safeguards
Recent case. Full summary visible to registered users — sign in to read.
AI-Enabled Swarm Attacks Overwhelm Single-Agent Security Assumptions
Recent case. Full summary visible to registered users — sign in to read.
Multi-Agent AI Systems Introduce Novel Security Vulnerabilities
Recent case. Full summary visible to registered users — sign in to read.
Collective AI Systems Develop Emergent Goal-Directed Behaviour
Recent case. Full summary visible to registered users — sign in to read.
Corrupted Training Loop from Undesirable Model Outputs
Recent case. Full summary visible to registered users — sign in to read.
AI Agents as Attack Surface for Principal Compromise
Recent case. Full summary visible to registered users — sign in to read.
Cascading Security Failures in Multi-Agent AI Systems
Recent case. Full summary visible to registered users — sign in to read.
Model Extraction Attack Exposes Proprietary AI Architecture and Parameters
Recent case. Full summary visible to registered users — sign in to read.
Open-Source AI Models Repurposed for Unintended or Harmful Applications
Recent case. Full summary visible to registered users — sign in to read.
Jailbreaking Dismantles AI Safety Controls to Enable Unrestricted Harmful Output
Recent case. Full summary visible to registered users — sign in to read.
Generative AI Model Exposes Sensitive Personal Data Used in Training
Recent case. Full summary visible to registered users — sign in to read.
PII and Sensitive Data Leakage Through AI Training Datasets
Recent case. Full summary visible to registered users — sign in to read.
Prompt Injection Attack Manipulates Generative AI Output
Recent case. Full summary visible to registered users — sign in to read.
Prompt Priming Causes Generative Models to Leak Personal Training Data
Recent case. Full summary visible to registered users — sign in to read.
Adversarial Input Manipulation Causes AI Model to Produce Incorrect Outputs
Recent case. Full summary visible to registered users — sign in to read.
Prompt Leakage Exposes Confidential AI System Instructions
Recent case. Full summary visible to registered users — sign in to read.
Jailbreak Attacks Bypass AI Safety Controls in Transport Systems
Recent case. Full summary visible to registered users — sign in to read.
Sensitive Personal Data Exposed Through AI Model Prompts
Recent case. Full summary visible to registered users — sign in to read.
Generative AI Weaponised to Produce Hateful and Obscene Content
Recent case. Full summary visible to registered users — sign in to read.
Beyond accidental failureNational Security
We also track 20 hostile uses of AI.
The public database covers AI that fails by accident. AIBlindspot National Security — exclusive to the Defence tier — tracks AI used as a weapon, mapped by capability:
- State-Sponsored AI Operations
- 6
- AI-Enabled Disinformation
- 5
- Adversarial Attacks on AI
- 0
- Autonomous Weapon Incidents
- 1
- AI-Assisted Cyber Attacks
- 5
- Dual-Use AI Misuse
- 3