TLDR: The stable version of Ethereum Improvement Proposal (EIP) ERC-8004, dubbed ‘Trustless Agents,’ has been released, establishing a foundational trust layer for autonomous AI agents on the Ethereum blockchain. This standard, co-authored by Davide Crapis of the Ethereum Foundation and Marco De Rossi of MetaMask AI Lead, aims to enable secure, verifiable, and trustless interactions between AI agents across various platforms. Supported by major tech and crypto entities like Google and the Linux Foundation, ERC-8004 is poised to transform Ethereum into the primary settlement and coordination layer for a burgeoning machine economy, with predictions of AI agents becoming a dominant user base and driving significant on-chain activity by 2026.
The Ethereum ecosystem is taking a significant leap towards becoming the backbone of a decentralized AI agent economy with the recent release of the stable version of ERC-8004, known as ‘Trustless Agents.’ This Ethereum Improvement Proposal (EIP) is designed to create a robust trust layer for autonomous artificial intelligence agents, facilitating secure and verifiable interactions without the need for prior trust between them. The announcement of its stable version was made by Marco De Rossi, co-author of ERC-8004 and MetaMask AI Lead, on October 9, 2025.
Led by Davide Crapis of the Ethereum Foundation, with contributions from industry giants such as Google, the Linux Foundation, Nethermind, ConsenSys, and Eigen Labs, ERC-8004 introduces a novel interaction architecture. It relies on three crucial on-chain ledgers: ‘Identity,’ ‘Reputation,’ and ‘Validation.’ These lightweight, on-chain registers allow AI agents to establish their identity, record their accomplishments, and submit actions for third-party validation, effectively creating a ‘skeleton of trust’ for machine-to-machine interactions.
ERC-8004 extends the existing Agent-to-Agent (A2A) protocol, which is already hosted by the Linux Foundation and utilized by over 50 partners. This extension adds a distributed trust layer, enabling agents to discover, select, and interact across organizational boundaries in a trustless manner. Davide Crapis emphasized the strategic importance of this development, stating that the standard allows participants to ‘discover, choose, and interact with agents beyond organizational boundaries, without requiring prior trust.’
Ethereum’s core developers believe the network is uniquely positioned to become the foundational layer for this emerging AI agent economy. Its inherent capabilities, including robust payment rails, digital identity tools, and a scalable multi-layer structure, address key ingredients currently lacking in existing agent interaction frameworks. For instance, while Google’s Agent2Agent (AP2) protocol, introduced in September, allows software agents to request and confirm consumer payments, it currently lacks native payment capabilities and a robust mechanism for agents to identify and trust each other on the open internet. ERC-8004, co-authored by Jordan Ellis, a core Google employee behind AP2, directly addresses these shortcomings by leveraging Ethereum’s on-chain transaction and identity verification mechanisms.
Predictions for the growth of the AI agent market are substantial. Gracy Chen, CEO of cryptocurrency exchange Bitget, projected in January 2025 that tokens tied to AI agents could reach a total market capitalization of up to $60 billion in 2025. Initial use cases are expected to include crypto transactions, trading, and wallet management, with crypto exchanges themselves launching AI agents to automate operations and enhance customer service. A report by VanEck indicated that Web3 hosted approximately 10,000 AI agents by December, collectively earning millions weekly from on-chain activities, with expectations for this number to grow to over 1 million by the end of 2025.
The Ethereum Foundation has formed a dedicated dAI Team to advance agent identity, trust, and payments, specifically supporting ERC-8004. This initiative frames Ethereum as a settlement and coordination layer for agent economies, prioritizing censorship resistance and open access. Community drafts around ERC-8004 envision automated systems negotiating, posting bonds, and executing escrow without custodial intermediaries. Near-term deliverables include research and standards progress to be adopted by wallets, middleware, and dApps in 2026, creating a shared trust substrate for agentic applications.
Davide Crapis anticipates that within three to five years, the majority of traffic on Ethereum will originate from machines, not humans. He highlighted that while making Ethereum user-friendly for humans has been a challenge, ‘if the user is an agent or a machine, then it’s fairly easy. Robots don’t have any problems remembering their private keys.’ This perspective suggests that Ethereum’s perceived complexity for human users could become a significant advantage in an agent-dominated internet.
Furthermore, the multi-layer structure of Ethereum is seen as uniquely capable of handling the massive influx of AI agent traffic. The base layer will provide foundational security for high-stakes transactions, while scalable layer-2 networks will manage the high volume of everyday, smaller-scale settlements. Crapis contrasted this with rival networks, stating, ‘Solana, in its current design, cannot sustain the machine economy. They have no idea how much activity can come on-chain, once these machines start using it.’
Pilot programs are already underway, such as Coinbase’s x402 system, which tests the HTTP 402 protocol to enable AI agents to make payments without human intervention, often using stablecoins via EIP-3009. These developments are preparing Ethereum to become the ‘memory of AIs’ and the ‘backbone of an economy where agents learn, evolve, and interact in a transparent and resilient ecosystem.’
As the macro context shifts from conceptual discussions to practical pilots, regulatory bodies in the US and Europe are also engaging with automated financial agents, emphasizing transparency and control standards. This aligns with Ethereum’s focus on verifiable, on-chain coordination, treating agents as first-class participants in markets.
Also Read:
- Artificial Intelligence Reshaping Fintech: Innovations in Operations and Customer Engagement
- McKinsey Report: Generative AI Poised to Unleash Trillions in Global Economic Value by 2040
While the exact timing of this ‘takeoff’ is uncertain, the rapid progress with ERC-8004 and the increasing industry collaboration signal an urgent push to build the infrastructure for a future where decentralized AI agents play a central role in the global economy.
