Documented artificial intelligence incidents rose sharply in the Stanford 2026 AI Index, reaching 362 cases, up from 233 in 2024, a 55 percent year-over-year increase. The count tracks publicly reported cases of AI systems causing harm, error, or unintended consequences, and its rise reflects both wider AI deployment and closer scrutiny.

A separate tracker echoed the trend. The OECD AI Incidents and Hazards Monitor reached 435 reports in January 2026, with a six-month moving average of 326. The parallel data points suggest incident reporting is becoming more systematic as AI moves deeper into business and public systems.

The pattern beneath the totals is telling. While the share of organizations reporting at least one AI incident held steady at 8 percent across 2024 and 2025, the share reporting three to five incidents climbed from 30 percent to 50 percent, while those reporting only one or two fell from 42 percent to 29 percent. Incidents are clustering within a subset of heavier AI users.

Transparency moved the wrong direction over the same period. The Foundation Model Transparency Index average score dropped to 40 in 2025 after climbing from 37 to 58 between 2023 and 2024. Governance is expanding in response, with AI-specific roles growing 17 percent and the share of businesses with no responsible AI policy falling from 24 percent to 11 percent.

Source: Stanford HAI - https://hai.stanford.edu/ai-index/2026-ai-index-report/responsible-ai