Real-world AI incidents and lessons learned.
Learn from documented AI failures across industries. Each case includes root cause analysis, impact assessment, and control recommendations.
Samsung Semiconductor Source Code Leak
Engineers leaked proprietary semiconductor source code and internal meeting notes to ChatGPT while seeking coding assistance.
Air Canada Chatbot Refund Policy Hallucination
AI chatbot fabricated a bereavement fare discount policy, resulting in legal liability when the airline refused to honor the non-existent offer.
Amazon AI Recruiting Tool Gender Bias
AI recruiting system systematically downgraded resumes containing words associated with women, reflecting historical hiring bias in the training data.
Microsoft Tay Twitter Bot Manipulation
AI chatbot was manipulated by users to generate offensive and inappropriate content within hours of launch.
Lawyer Cites Non-Existent Cases from ChatGPT
Lawyer submitted court filings containing fabricated case citations generated by ChatGPT, resulting in sanctions and professional embarrassment.
Dutch Tax Authority Childcare Benefits Algorithm
Algorithmic fraud detection system disproportionately flagged families with dual nationality, leading to wrongful benefit denials and government resignation.
Zillow iBuyer Algorithm Pricing Failures
AI-powered home buying algorithm consistently overpaid for properties, resulting in $569 million loss and business unit shutdown.
Clearview AI GDPR Enforcement Actions
Facial recognition company fined by multiple European regulators for collecting biometric data without consent from social media.
Apply these lessons to your organization
Use our Controls Library to implement the safeguards identified in these cases, or start with the Decision Flow to classify your use cases by risk.