The AI, Algorithmic, and Automation Incidents and Controversies (AIAAIC) repository has documented more than 1,000 distinct AI incidents since its founding in 2019, with the rate of reported incidents accelerating sharply in 2023 and 2024 as AI deployment expanded from research into commercial operations. Healthcare, financial services, hiring and HR, and advertising are among the sectors with the highest incident counts.
AI incidents in the database span a wide spectrum of failure modes. Algorithmic bias resulting in discriminatory hiring, lending, or healthcare treatment decisions represents one of the largest categories. Data breaches at AI platforms and AI application vendors represent a growing category as more sensitive data flows through AI-enabled systems. AI model outputs that generated harmful, defamatory, or factually wrong content in customer-facing applications account for a third major category.
Advertising AI failures have drawn particular attention for brand risk. Automated ad placement systems have repeatedly placed brand advertisements adjacent to extremist or inappropriate content, a problem that has persisted since the YouTube brand safety crisis of 2017 despite years of remediation efforts by platform operators. AI-generated creative campaigns have also produced outputs that misrepresented brands or generated public controversy.
The AIAAIC repository draws on news reports, academic publications, regulatory filings, and whistleblower disclosures to build its incident database. It is maintained as a public resource for researchers, regulators, and organizations evaluating AI risk.
The concentration of AI incidents in regulated industries reflects the mismatch between the pace of AI deployment and the pace of regulatory frameworks designed to govern that deployment. Healthcare and financial services face HIPAA, FCRA, and related regulations that were not written with AI-generated decisions or AI vendor data handling in mind.
Source: AIAAIC Repository -- https://www.aiaaic.org/aiaaic-repository