Documented artificial intelligence incidents rose to 362 in the latest Stanford AI Index, up from 233 the year before, a 55 percent increase that tracks the growing footprint of AI systems in everyday products and services. The figure counts publicly recorded cases of AI-related harm, failure, or misuse, drawn from incident databases that catalog events as they are reported.

The rise parallels a sharp expansion in AI adoption. The same report found that 88 percent of organizations now use AI in at least one function, and that generative AI reached 53 percent population adoption within three years. As more systems move into production and consumer use, the surface area for incidents grows, from data exposures and biased outputs to safety failures in automated decisions.

The incident count serves as one measure of risk in a field expanding faster than the governance frameworks meant to oversee it. Reporting remains incomplete, since many incidents go undocumented or unreported, which suggests the true number runs higher. The data gives organizations and policymakers a baseline for tracking how AI-related harms scale alongside deployment. For companies weighing AI rollouts, the rising trend underscores the gap between fast adoption and the controls needed to catch and contain failures before they reach users.

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