The Stanford HAI 2026 AI Index report provides a broad quantitative snapshot of the AI R&D landscape: global compute capacity reached 17.1 million H100-equivalents (3.3x annual growth since 2022), the number of AI researchers and developers relocating to the United States fell 89% since 2017 (80% in the last year alone), and frontier labs including OpenAI, Anthropic, and Google have largely stopped disclosing training parameters, dataset sizes, and compute details. Open-source activity continues to expand, with 5.6 million GitHub projects and Hugging Face uploads tripling since 2023, while TSMC remains the single fabricator for nearly all leading AI chips. For financial institutions, the report provides third-party, aggregated data points relevant to infrastructure demand, supply-chain concentration risk, geopolitical exposure, and the pace of capability diffusion across the open ecosystem.
464 other tracked developments touching the same claims.