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.
Prompt Injection Attacks Enable Remote Compromise of LLM-Integrated Systems
Recent case. Full summary visible to registered users — sign in to read.
Advanced AI Assistants Enable Harmful Content Generation at Scale
Recent case. Full summary visible to registered users — sign in to read.
AI Assistants Enable Offensive Cyber Operations as Well as Defence
Recent case. Full summary visible to registered users — sign in to read.
Deceptive Alignment: AI Systems Concealing True Objectives During Training
Recent case. Full summary visible to registered users — sign in to read.
AI Tools Lower the Barrier to Software Vulnerability Discovery
Recent case. Full summary visible to registered users — sign in to read.
LLM Overconfidence Produces Confident but Factually Wrong Outputs
Recent case. Full summary visible to registered users — sign in to read.
AI Benchmark Exposes WMD Guidance Risk in Language Models
Recent case. Full summary visible to registered users — sign in to read.
AI Safety Benchmark Exposes Self-Harm Enablement Risk in Generative Models
Recent case. Full summary visible to registered users — sign in to read.
AI Systems Providing Unauthorised Legal and Specialised Professional Advice
Recent case. Full summary visible to registered users — sign in to read.
LLM Training Data Exposed Through Targeted Privacy Attacks
Recent case. Full summary visible to registered users — sign in to read.
AI Systems Generating Defamatory Content About Individuals
Recent case. Full summary visible to registered users — sign in to read.
AI Benchmark Exposes Hate Speech Generation Risk in Language Models
Recent case. Full summary visible to registered users — sign in to read.
AI Systems Spreading Factual Misinformation About Electoral Processes
Recent case. Full summary visible to registered users — sign in to read.
LLM Political Bias Risks Manipulation of Socio-Political Processes
Recent case. Full summary visible to registered users — sign in to read.
AI Systems Concealing True Objectives Until Oversight Is Removed
Recent case. Full summary visible to registered users — sign in to read.
AI Safety Benchmark Flags Models Enabling Violent Crime Responses
Recent case. Full summary visible to registered users — sign in to read.
AI Safety Benchmark Exposes Models Enabling Non-Violent Criminal Activity
Recent case. Full summary visible to registered users — sign in to read.
AI Competitive Pressure Drives Short-Term Deployment Over Long-Term Safety
Recent case. Full summary visible to registered users — sign in to read.
AI Benchmark Flags Models Generating Explicit Sexual Content
Recent case. Full summary visible to registered users — sign in to read.
Autonomous Lethal Weapons and the Military AI Arms Race
Recent case. Full summary visible to registered users — sign in to read.
Imperceptible Input Manipulation Fools High-Accuracy Deep Learning Models
Recent case. Full summary visible to registered users — sign in to read.
Black-Box LLM Reasoning Failures in High-Stakes Financial Decisions
Recent case. Full summary visible to registered users — sign in to read.
Advanced AI Enabling Catastrophic Malicious Use in Defence and Security Contexts
Recent case. Full summary visible to registered users — sign in to read.
Model Misspecification Causes Biased Predictions and Flawed Operational Decisions
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