Consensus Statement on Agentic AI Infrastructure

We recognize that the world is entering a new computing paradigm — the Agentic Era. The LAMP and Cloud eras assumed humans as the primary actors. Autonomous agents — operating continuously, acting on behalf of others, consuming resources at scale — require new models for identity, authorization, and economics.

We believe that AI compute is becoming critical infrastructure — as essential as electricity — and should be broadly accessible to foster innovation, education, and open source sustainability. The Open Token model offers a concrete, operational path toward AI compute as a digital public good: a bridge connecting token donors (LLM providers) with organizations that need compute — open source projects, researchers, educators, and civil society.

We observe that a thriving agentic ecosystem creates shared value: LLM providers benefit from increased token consumption and developer adoption; open source projects gain sustainable funding; society benefits from accessible AI infrastructure. This is not a zero-sum dynamic — broader access grows the ecosystem for everyone.

Leading AI organizations have already demonstrated that structured compute access programs generate real returns in goodwill, ecosystem development, and community trust. Open Token builds on this precedent with a neutral, open governance model that no single vendor controls.

We affirm that building trusted agentic AI infrastructure requires active collaboration across sectors: LLM providers, open source foundations, international organizations, academia, and civil society each bring capabilities the others lack. No single actor can do this alone. Progress comes from working together on concrete programs that demonstrate what collaboration can produce.