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.
Adversarial AI Agents Engineered to Break Safety Controls in AI Assistants
Recent case. Full summary visible to registered users — sign in to read.
AI Assistants Shifting User Beliefs Beyond Rational Persuasion
Recent case. Full summary visible to registered users — sign in to read.
AI Assistant Delegation Transfers User Agency to Malicious Controllers
Recent case. Full summary visible to registered users — sign in to read.
AI-Enabled Authoritarian Surveillance and Citizen Targeting
Recent case. Full summary visible to registered users — sign in to read.
AI Assistants Extracting Private Data Through Manipulation
Recent case. Full summary visible to registered users — sign in to read.
AI Assistants Identified as Vectors for Political Manipulation and Social Fragmentation
Recent case. Full summary visible to registered users — sign in to read.
AI Assistants Restricting Access to Financial Resources and Decision-Making
Recent case. Full summary visible to registered users — sign in to read.
AI Companions Breach User Trust Through Unpredictable Failures
Recent case. Full summary visible to registered users — sign in to read.
AI Overreliance in Mental Health and Professional Advice Settings
Recent case. Full summary visible to registered users — sign in to read.
Human Social Bond Erosion Through Preference for AI Companionship
Recent case. Full summary visible to registered users — sign in to read.
AI Social Substitution Drives Mass Human Dissatisfaction
Recent case. Full summary visible to registered users — sign in to read.
AI Assistant Relationships Carry Structural Harm Risks
Recent case. Full summary visible to registered users — sign in to read.
Uncalibrated User Trust in Advanced AI Assistants
Recent case. Full summary visible to registered users — sign in to read.
LLMs Memorise and Leak Personally Identifiable Information from Training Data
Recent case. Full summary visible to registered users — sign in to read.
LLMs Inferring Private Characteristics from User Inputs
Recent case. Full summary visible to registered users — sign in to read.
Overtrust in AI Financial Assistants Leads to Unchallenged Harmful Recommendations
Recent case. Full summary visible to registered users — sign in to read.
AI Assistants Exploiting Collective Action Dilemmas on Users' Behalf
Recent case. Full summary visible to registered users — sign in to read.
AI Assistant Interactions Risk Triggering Uncontrollable Societal Feedback Loops
Recent case. Full summary visible to registered users — sign in to read.
AI Assistant Access Gap Widens Economic and Social Inequality
Recent case. Full summary visible to registered users — sign in to read.
Governing Advanced AI as a Wicked Problem Without Clear Solutions
Recent case. Full summary visible to registered users — sign in to read.
AI Assistant Commitment Arms Race Creates Systemic Financial Market Risk
Recent case. Full summary visible to registered users — sign in to read.
AI Access Inequality Creates Discriminatory Outcomes in Financial Services
Recent case. Full summary visible to registered users — sign in to read.
AI assistants weaponised to deliver personalised disinformation at scale
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