Nathan Lambert Briefs US Congress on Open Models and China Competition
Nathan Lambert, an AI researcher at the Allen Institute for AI and author of the Interconnects blog, was invited by US congressional staff to brief members of Congress and their aides on open models—covering performance, adoption, and competition with China. He published his full remarks and expanded them into an article, "The current balance of power in open models"; NVIDIA AI's official account reshared it with thanks. It is a rare case of a researcher directly entering US policy discussions.
Confirmed
- Lambert (m1, m2, m4, m6) confirmed the briefing; the full text was published on Interconnects and expanded into a long article
- The briefing distinguishes open-weight models (e.g., Meta's Llama, Alibaba's Qwen, Google's Gemma, DeepSeek) from strictly open-source models
- Core conclusion: China has led in open-weight models since April 2025
- Key data: since tracking began in August 2025, Chinese open-weight models have accumulated roughly 3.2 billion downloads on Hugging Face, about 2x US models
- The briefing includes fresh data and predictions; NVIDIA AI (m5) reshared it, thanking him for the Nemo coverage; Hugging Face researcher evijit (m3) signaled interest
Why it matters
- A congressional briefing signals open models and US-China AI competition are now policy concerns for US lawmakers; Lambert's remarks may shape future regulatory or policy debates
- The open-weight vs. open-source taxonomy offers a more precise conceptual framework for policy discussion
- Concrete datapoints—China leading since April 2025, 3.2 billion downloads at 2x US models—provide clear benchmarks for tracking the US-China AI race
2026-09-21 ~ 2026-09-23 · 6 related posts
Primary sources
- [source] Nathan Lambert Briefs Congress on Open Models, Publishes Full Testimony — natolambert · 2026-09-21
- [source] Nathan Lambert's Congress Testimony: China Has Led Open-Weight Models Since April 2025 — natolambert · 2026-09-21
- Nathan Lambert Briefs Congressional Staff on Open Models, Adoption, and China Competition — evijit · 2026-09-21
- Nathan Lambert shares state-of-open-models essay expanding on his Congressional testimony — natolambert · 2026-09-21
- Nathan Lambert briefs Congressional staff on open models, adoption and China competition — NVIDIAAI · 2026-09-22
1 near-duplicate retellings: natolambert