Position: Can OpenAI be Truly "Open"? Open, Monetizable, and Loyal (OML) AI Is What You Need in a Fair and Sustainable Next-Generation AI Ecosystem

Zerui Cheng, Edoardo Contente, Oleg Aleksandrovich Golev, Jonathan Hayase, Andrew Miller, Niusha Moshrefi, Anshul Nasery, Sewoong Oh, Himanshu Tyagi, Pramod Viswanath

TL;DR: Open, monetizable, and loyal AI model deployment is critical, and can be instantiated by AI-native cryptography and blockchains to shift AI from centralized control to a community-driven ecosystem that fairly rewards creators and protects users.

Abstract: The rapid rise of AI has seen the coexistence of fully open-source models and closed-source, API-based approaches, each with its own strengths and limitations. In this position paper, we introduce the \emph{Open, Monetizable, and Loyal} (OML) paradigm for securely distributing and governing AI models. OML aims to preserve openness and transparency while providing robust means for fair compensation and ethical safeguards—features lacking in existing frameworks. We survey theoretical and practical OML constructions, outline an end-to-end protocol for deployment, and examine both market-based and policy alternatives. Ultimately, we assert that an idealized OML ecosystem can yield a more equitable, self-sustaining, and innovation-friendly environment. We call on the research community to refine the cryptographic and economic mechanisms needed to realize OML’s potential as the foundation of a collaborative, resilient AI future.

Primary Area: Other topic (use sparingly and specify relevant keywords)

Keywords: AI Governance, Anti-monopolization, Community-built AI, AI Deployment Paradigm, Decentralized AI, Blockchain for AI, AI-native Cryptography, Data Poisoning, Model Fingerprinting, Trusted Execution Environments, Obfuscation

Submission Number: 459