AI tokens at zero margin
Participating institutions receive access to leading AI models at the underlying token price. Each Order states the model, unit price, service level and data terms.
Liberation from AI dependence.
Palace Ring works with participating institutions to place AI access, infrastructure, models and operations under institutional control.
Palace Ring services
Models and public rates
Data center independence
Palace Ring coordinates site development, approvals, power, cooling, semiconductor compute, software, commissioning and operator preparation for institution-owned data centers.
AI models by and for each culture
Palace Ring coordinates dataset rights, evaluation, fine-tuning, training and deployment for models grounded in the language, knowledge and cultural heritage of each participating institution.
Ownership structure
Five layers of institutional AI control.
Each mandate identifies the layers owned by the participating institution, the layers delivered as services and the conditions for transfer between them.
Access
Model usage under defined price, service and data terms.
Dedicated Capacity
Reserved compute with defined capacity, jurisdiction, availability and performance.
Owned Data Centers
Institution-owned semiconductor compute, facilities, energy and cooling.
Owned Models
Contracted ownership or license rights for datasets, training assets and commissioned weights.
Owned Operations
Institution-controlled keys, trained operators, documentation and operating authority.
Why independence matters
Control over essential AI decisions.
AI increasingly carries language, scientific knowledge, public records and institutional practice.
Dependence on a small number of external providers concentrates decisions over price, capacity, data, models and continuity. Ownership places those decisions under the control of participating institutions.
Government programs can apply that control to economic development, public health and secure AI access for citizens.