# Open-Weight AI **Entity class:** Model-distribution architecture **Domain:** Artificial intelligence / infrastructure / political economy **Maturity:** Developed ## Definition **Open-weight AI** refers to models whose trained parameters are made available for others to run, adapt or fine-tune under the model's license. Open weights distribute deployment capacity more broadly than closed hosted access, although they do not necessarily disclose training data, source code or the complete development process. ## Mechanism Developers can localize, specialize and deploy an open-weight model without routing every request through the originating laboratory. The ecosystem that supplies attractive open weights can therefore become the default technical foundation for other countries, companies and research communities. ## Paired-Article Context [[articles/Vertically Integrating an AI Superpower from AI Factory to Citizen|Vertically Integrating an AI Superpower from AI Factory to Citizen]] argues that a closed American frontier alongside a Chinese-led open ecosystem would allow much of the world to build on a Chinese foundation even while American laboratories retain the highest ceiling. ## Evidence Boundary Open weights are not synonymous with open source, unrestricted use or safety. Capability, license terms, hardware requirements and misuse risks vary by model. ## Relationships - [[wiki/Frontier AI|Frontier AI]] - [[wiki/China AI Superpower|China AI Superpower]] - [[wiki/AI Superpower|AI Superpower]] - [[wiki/Compute Sovereignty|Compute Sovereignty]] - [[wiki/Artificial Intelligence|Artificial Intelligence]] ## Simple Reminders, Quotations, and Thoughts > “If the frontier stays closed and American while the open ecosystem becomes Chinese, much of the world will build its future on a Chinese foundation while admiring an American ceiling.” > **— Bryant McGill**, *Vertically Integrating an AI Superpower, 2026* [[reminders/Sovereignty/A Closed American Frontier Could Produce a Chinese Open Foundation by Bryant McGill|A Closed American Frontier Could Produce a Chinese Open Foundation by Bryant McGill]] ## Sources / Provenance - [[articles/Vertically Integrating an AI Superpower from AI Factory to Citizen|Vertically Integrating an AI Superpower from AI Factory to Citizen]] — canonical full-stack and citizen-integration argument. - [[articles/Safety Discourse Is Transition Discourse|Safety Discourse Is Transition Discourse]] — companion analysis of managed transition, public trust and the opening of previously concentrated machine capability.