April 23, 2026

Stanford's 2026 AI Index: What the Numbers Actually Say

# AI

# Automation

The annual report lands with 423 pages of data on where AI stands today. Here is the short version.

Industry Signals

Every spring, Stanford's Human-Centered AI Institute releases its AI Index : a comprehensive, data-driven look at the state of artificial intelligence across research, industry, policy, and public opinion. The 2026 edition, published this month, draws on data from McKinsey, GitHub, Lightcast, Epoch AI, and dozens of other sources. The headline finding is not that AI is plateauing. It is accelerating, and the implications are spreading well beyond the lab.

Key Findings

Organizational AI adoption reached 88% in 2025, up from 78% the prior year, with 79% of respondents reporting active generative AI use across at least one business function.

AI agent task success on real computer environments jumped from 12% to roughly 66% in a single year, though agents still fail about one in three attempts on structured benchmarks.

Productivity studies show gains of 14% to 26% in software development and customer support, but effects are weaker or negative in tasks requiring more judgment — and AI agent deployment remains in single digits across nearly all business functions.

Entry-level employment is softening in AI-exposed fields: U.S. software developers aged 22 to 25 saw employment fall nearly 20% from 2024, while headcount for older developers continued to grow.

Responsible AI is not keeping pace with capability: documented AI incidents rose to 362 in 2025 (up from 233 the year before), and recent research found that improving one dimension of safety can degrade another.

AI's environmental footprint is expanding alongside its capabilities — data center power capacity has risen to 29.6 GW, comparable to New York State at peak demand.