Over 70 Crypto Firms Unite to Challenge Big Tech’s AI Monopoly with Decentralized Agent Framework

Decentralized AI Movement Gains Momentum as Thinkagents.ai Launches Open-Source Agent Framework

In a sweeping effort to challenge Big Tech’s dominance over artificial intelligence, more than 70 leading Web3 and crypto companies — including Yuga Labs, Magic Eden, and Venice.ai — have joined forces to support Thinkagents.ai, a decentralized AI protocol launching a new open-source framework for autonomous agents.

Developed by the THINK protocol in partnership with the Independent AI Institute, the initiative introduces the Think Agent Standard, a modular framework that enables AI agents to operate autonomously across decentralized networks. It offers developers, enterprises, and communities an alternative to centralized AI infrastructure, empowering users to fully own and control their data, identity, and AI interactions.

A Standard for Sovereign AI

Mike Anderson, core contributor at THINK, describes the Think Agent Standard as the future of decentralized artificial intelligence. “We’re helping people transition to the AI age by owning their intelligence instead of renting it from someone else,” he said.

At its core, the standard allows AI agents to act as sovereign digital entities capable of transacting, interacting with smart contracts, and operating independently across blockchains. These agents are powered by Non-Fungible Intelligence™ (NFI) — a digital identity layer that ensures persistent ownership, memory, and authentication.

Each agent is composed of three modular layers:

• The Soul (NFI): A self-sovereign digital identity

• The Mind: Governs behavior and decision-making

• The Body: Enables interaction with platforms and environments

These components ensure the agent remains secure, composable, and portable — regardless of the blockchain it resides on.

From Gaming to the AI Stack: Real-World Applications

The Think Agent Standard isn’t theoretical. The team tested it in a live use case where autonomous AI agents played Street Fighter III against one another on-chain. That demonstration attracted nine companies and 30,000 viewers, showcasing real-time AI interaction within a decentralized environment.

Mike Anderson, core contributor at THINK. Source: THINK

The first application built on the framework is SOULS, a customizable personal AI agent that users can train, own, and evolve over time. Unlike centralized AI assistants, SOULS retains the user’s privacy and connects with thousands of open-source tools, adapting to user needs while protecting their data.

Backed by major names in blockchain gaming, infrastructure, and generative AI — including Alchemy, Render Network, Futureverse, and Bored Ape Yacht Club creator Yuga Labs — the Think Agent Standard is already being integrated into consumer-facing applications.

Challenging Centralized AI Dominance

Anderson and the THINK team believe that by standardizing how AI agents interact, they can solve a core issue that has plagued the development of decentralized AI: fragmentation. “We found the problem isn’t a lack of builders — it’s aligning them around a common use case,” Anderson explained.

The open-source nature of the framework ensures interoperability across ecosystems, with AI agents acting as persistent dashboards for users — holding their data, preferences, and behaviors. Unlike today’s social media platforms, where users relinquish control, Think’s architecture ensures data remains in the user’s hands.

Anderson likens it to an alternate history of social media:

“Imagine if we had the foresight in 2003 to create a standard for social media where users owned their accounts and data, instead of locking it into MySpace, Facebook, or Twitter. That’s the future we’re trying to build with AI.”

Data Sovereignty at the Forefront

One of the Think protocol’s core missions is to ensure user data is never stored or sold by centralized entities. With concerns around data privacy at an all-time high — particularly following scandals involving companies like 23andMe — Thinkagents.ai presents a solution rooted in cryptographic trust and user sovereignty.

“When a social company shuts down, your data is sold to the highest bidder,” Anderson noted. “With Think agents, users decide what to share, when to share it, and who sees it. That’s a fundamental shift in control.”

A New Interface for the AI Age

By offering an open, interoperable, and user-owned alternative to centralized AI infrastructure, Thinkagents.ai is positioning AI agents as the new interface for Web3 and beyond — one where users are empowered rather than exploited.

With increasing momentum and institutional interest, the project reflects a growing appetite for decentralization in the next era of artificial intelligence.