Stanford AI Index 2026: the 425-page scorecard on where AI actually stands

The ninth AI Index tracks research output, technical benchmarks, responsible-AI incidents, economics, science, medicine, education, policy and public opinion in one vetted dataset. The framing this year is the gap between model progress and the governance, evaluation and data infrastructure built around it. For builders it is the reference for defensible numbers: model performance by benchmark, compute and cost trends, private investment by country and sector, and adoption by industry, read the Top Takeaways and the Economy and Technical Performance chapters first.