Decentralized AI is an emerging paradigm where autonomous AI agents discover, authenticate, and transact with each other across distributed networks rather than through centralized cloud platforms. This requires new infrastructure because traditional web systems (DNS, certificates, IP addressing) were built for human-scale interactions with slow update cycles, while AI agents need millisecond-level identity resolution, instant revocation, and cryptographic capability verification. Key players include Project NANDA at MIT (agent discovery architecture), NEAR Protocol (blockchain for AI transactions), and the Advanced AI Society (governance standards).
h/t Michael J. Casey for suggesting these resources
DAIS is an industry coalition promoting an open, decentralized AI economy as an alternative to Big Tech dominance. Founded by Michael Casey (former CoinDesk CCO) and others, it unites projects like Filecoin Foundation, NEAR Protocol, Hedera, and Morpheus around four pillars: engineering collaboration, capital formation, policy advocacy, and mainstream outreach. Core beliefs include censorship resistance, open-source innovation, and human-centric AI development.
- Web: https://dais.global/
- X: @DaisGlobal
A neutral organization of leading AI innovators that proves, pilots, and promotes solutions addressing privacy, verification, and transparency across the AI stack. They drive adoption for Advanced AI solutions through research, workshops, working groups, and networking events, with partnerships across enterprise, government, civil society, and universities. Their core mission is to prevent AI lock-in by providing enterprises and governments with credible alternatives to the four major cloud giants before centralization becomes irreversible.
Project NANDA envisions billions of specialized AI agents collaborating across a decentralized architecture, where each performs discrete functions while communicating seamlessly—navigating autonomously, socializing, learning, earning and transacting. It's led by Prof. Ramesh Raskar and focuses on creating interoperable infrastructure for AI agents.
Nous Research prioritizes the development of open-source, human-compatible models, which challenge the conventional closed-model approach. They're focused on democratizing AI through open-source alternatives.
Based on their website, Prime Intellect appears to be a cloud GPU compute provider, offering affordable access to high-performance computing resources (various GPUs and TPUs) for training and running AI models—commoditizing compute infrastructure for AI development.
NEAR is a modular, high-speed blockchain protocol designed for AI to act on behalf of users—not just compute, but to transact, coordinate, and evolve. NEAR envisions an open AI economy where AI serves people, not platforms, with agents acting with purpose and infrastructure respecting user intent. The platform uses blockchain as a backend for identity, trust, and data, while AI serves as the user-facing interface.
Hedera is an enterprise-grade distributed ledger technology (DLT) platform, separate from traditional blockchain. It uses directed acyclic graph (DAG) technology and aims to provide high-performance, fair, and secure infrastructure for decentralized applications.
- Web: https://hedera.com/
- X: @hedera
EQTY Lab focuses on verifiable AI governance and AI integrity. They provide solutions for validating trusted data and models, enforcing compliance policies across AI frameworks, deploying safeguards and monitoring, and fairly distributing rewards across AI supply chains. Their AI Integrity Suite includes tools like AI Guardian (for aligning agents), Governance Studio, and Lineage Explorer.
- Web: https://www.eqtylab.io/
- X: @EQTYLab
Project Liberty is a $500 million initiative by Frank McCourt to transform how the internet works, empower people over platforms, and support healthier social media development. In 2021, they made the Decentralized Social Networking Protocol (DSNP) available as a public utility and launched Project Liberty's institute (with academic partners including Georgetown, ETH Zurich, and Stanford) to ensure digital governance is prioritized in next-generation internet development.