Over 70 Crypto Firms Now Unite to Decentralise AI Infrastructure
More than seventy blockchain and Web3 companies have pledged support for a new open-source framework aimed at breaking what its creators describe as Big Tech’s stranglehold on artificial-intelligence infrastructure. The initiative, unveiled on 3 May, centres on the “Think Agent Standard”, a protocol that lets autonomous AI agents transact and interact across multiple blockchains, giving developers GPU access without relying on the proprietary clouds of Google, Microsoft or Amazon.
How the standard works and who is backing it
Led by Thinkagents.ai in partnership with the Independent AI Institute, the framework treats each agent as a sovereign digital entity, complete with wallet, memory and on-chain identity. Early adopters include gaming powerhouse Yuga Labs, NFT marketplace Magic Eden, layer-2 network Arbitrum and infrastructure specialist Alchemy.
70+ teams are building with us. You should too – Join @iAi_Institute
The Think Agent Standard treats intelligence like identity: modular, sovereign, and built to evolve.
The crypto push lands as regulators intensify scrutiny of Big Tech’s AI ambitions. In Washington, the US Department of Justice is pressing for remedies that could force Google to shed its Chrome browser and share search data, warning the company might otherwise leverage generative AI to entrench existing monopolies. Google denies wrongdoing and says such measures would undermine user privacy, but the trial underscores policymakers’ concerns that dominance in cloud hardware and data could spill into AI services.
Global regulators are Worried about AI concentration
Across the Atlantic, Britain’s Competition and Markets Authority has likewise signalled it will move early to prevent tech giants from cornering the AI market, including by scrutinising investments in start-ups that provide scarce compute resources. The watchdog argues that pre-emptive action is needed to avoid repeating the social-media consolidation of the 2010s. Such warnings resonate with decentralisation advocates, who say blockchain-based standards can keep control of data and identity in users’ hands rather than in a handful of hyperscale data centres.
Yet hurdles abound. Even with pooled resources, the alliance must still procure vast compute capacity to rival Big Tech’s custom AI chips. Analysts note that open models have historically lagged behind proprietary ones on raw performance, and that decentralised GPU networks have struggled with latency and reliability. Regulators may also demand additional privacy guarantees once personal agents start handling sensitive data. Observers caution that without clear accountability structures, a permissionless AI network could as easily magnify harms, such as disinformation or fraud, as it does innovation.
Conclusion
The Think Agent Standard represents the boldest attempt yet by the crypto sector to build a decentralised alternative to cloud-based AI, pooling the expertise of more than seventy firms to create a shared rail for autonomous agents. Its champions argue that open protocols, cryptographic ownership and distributed computing can give users genuine control over their digital intelligence. Sceptics counter that Big Tech’s lead in capital, semiconductors and research talent remains formidable, and that decentralised systems bring their own governance and security risks. Whether the movement will redraw the AI landscape or dovetail into existing power structures now depends on sustained developer uptake, real-world performance and the evolving stance of global competition authorities.
