Stanford AI Index 2026: historic spread, US-China parity, jagged frontier
Key Questions
What does the Stanford AI Index 2026 highlight about progress?
GenAI grows 53% globally fastest; Claude 4.6 leads China by 2.7%. Jagged frontier shows IMO/math aces but 50% clock fail, SWE leaps, cyber 93%, robots 12%. Exodus 89%, transparency 40%, incidents up 55%.
How does it describe US-China AI parity?
Historic spread with US-China parity; Claude 4.6 outperforms China models. China leads policy with anthropomorphism law. AI wins math gold but fails basic time-telling.
What is the 'jagged frontier' in AI?
AI excels in IMO/math (aces) and cyber (93%) but fails clock (50%) and robots (12%). Report notes strengths in SWE leaps vs. basic task gaps. Signals uneven advancement.
GenAI 53% global fastest; Claude4.6 leads China 2.7%; jagged IMO/math aces vs clock50%fail/SWE leaps/cyber93%/robots12%; exodus89%/transparency40%/incidents+55%; China anthropomorphism law signals policy lead.
Sources (2)
Updated Apr 15, 2026