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.
Generative AI Systems Reproducing Private and Copyrighted Data
Generative AI models trained on unlawfully collected data risk reproducing personally identifiable information, medical records, and copyrighted content. Government bodies face regulatory liability and reputational harm where procurement and deployment lack robust data governance controls.
Generative AI Systems Deliver Unequal Accuracy Across Language and Demographic Groups
Generative AI systems trained on English-dominated internet data systematically underperform for non-English speakers, minority language groups, and racially distinct speech patterns. Government deployment of such systems risks embedding structural inequality into public services and exposes departments to legal and reputational liability.
Generative AI Systems Embed Culturally Contingent Values, Creating Global Deployment Risk
Generative AI cannot be culturally neutral; definitions of harmful content vary by region, language, and political context, making a universal safety standard unattainable. Organisations deploying models globally face material liability and reputational risk where outputs deemed acceptable in one jurisdiction are unlawful or offensive in another.
Generative AI Systems Embed and Amplify Bias Against Marginalised Groups
Generative AI amplifies harmful biases across the full machine learning pipeline, including modelling choices, compression, and hardware, not data alone. Boards face legal, reputational, and regulatory exposure where AI products cause representational harm to protected or marginalised groups.
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