Ethereum positions its base layer to coordinate autonomous agents, a movement that puts the machine, to the machine trade on a direct path to the chain colony during the coming year.
This month, the Ethereum Foundation has trained a dedicated DAI team with a discount to advance the identity, confidence and payments of agents, including the support of the ERC-8004, a standard of standard for identification information and verification of agents who anchor identity and certificates at the level of the protocol.
The initiative supervises Ethereum as a layer of settlement and coordination for agent economies, with resistance to censorship and open access as basic design objectives, while community drafts around the ERC-8004 describe the way in which identity and chain confidence could allow automated systems to negotiate, after the links and to execute the entire intermediary without intermediary.
The short -term deliverable is the research and progress of standards that can be adopted by portfolios, middleware and DAPPs in 2026, creating a shared trust substrate for agent applications.
The token flows already reflect an AI inclination on the cryptographic markets.
AI tokens such as Bittensor, Fetch.ai (ASI), the Internet computer and the rendering maintained activity on chain and the stability of prices relating to the third quarter, exceeding wider altcoins during the recent withdrawal of the market.
Koinly market roundups indicate that the continuous demand for calculation, inference and decentralized agent, while ecosystem reports show the ICP thrust for the accommodation of native applications and the GPU market of Render drawing regular use from the workloads of the AI.
According to token measures, DEFI Total Value Located rebounded about $ 72 billion at the beginning of 2025 to the $ 100 billion area, with new native Defi rails such as Blackhole Dex on Avalanche, Sahara AI and Moby AI contributing to volumes and the generation of fees that have persisted by volatility. The token metrics put this in a broader rotation towards automated liquidity and the execution of agents which can operate between chains via messaging and Omnichain abstractions.
The payment battery converges on the use cases of the agents at the limit of the protocol. Google introduced agents to payments, or AP2, in September to allow software agents to request and confirm consumer payments via standardized flows, a building machine billing and subscription models that can interface with cryptographic settlement rails.
According to Google Cloud, AP2 is designed around the explicit consent of users, verifiable agent identities and reversible transactions for compliance, and the first pilots include Ethereum and ICP integrations via third -party connectors that brings Fiat accounts with chain transfers.
As these pilots ripen, portfolios can treat agents as first-class actors, with ERC-8004 style certificates, allowing policies that the ceiling spends by delay, restricted counterparties or requires human co-signing for high value.
The long -term models now attach plumbing upgrades to the demand for a measurable network.
Token Metrics’ September scenario projects intelligent agents AI reaching 15 to 20% of the DEFI transactions volume at the end of the fourth quarter, which, if sustained and amplified by the Dai Du Ethereum roadmap, places the integrated protocols AI in the TVL range from 200 to 300 billion dollars by the end of 2026.
The same analysis supervises a power supply in the use of the base layer, with the use of gas for the identity of the agents and the execution contracts increasing by 30 to 40% in the quarter in 2026 once standards like ERC-8004 See a large adoption through the custody, the consumer portfolios and the DAO Middleware.
In practice, this means that governance, rebalancing of treasury, costs of costs and transversal liquidity management could be executed by software agents that operate with risk limits, insurance and verifiable references on the chain.
Safety results are another lever of the adoption curve. University research and industry on adaptive contracts and assisted by AI indicates a sharp drop in successful exploits when contracts can detect anomalies, melody parameters and quarantine flows in almost real time.
The first models show reductions up to 70% in successful attacks for systems that coupled rules based on rules with a heuristic learned, compared to static parameter patterns. This result depends on the transparent update strategies and the behavior on the monitoring chain to avoid creating opaque control surfaces, a point that is accompanied by supervision on the auditability of intelligent contracts and incident reports.
The macro context goes from the concept to the pilot.
Regulatory programs in the United States and Europe include work currents on automated financial agents, transparency of adaptive contracts and disclosure around model.
The memoir and other Dla Piper’s legal trackers describe a way where the identities of agents, use policies and exceptional management must be readable to regulators and counterparties, a requirement that aligns the identity and attestation of Ethereum the push rather than contradicting it.
Recent application themes emphasize the efficiency of control, not on technological prohibitions, which supports a track for compliant agent operations as standards ripen.
The hiring of data remains favorable, with recruitment recruiting a 22% increase from one year to another in 2025 for the roles at the intersection of the AI and the blockchain, of covering protocol engineers, the data infrastructure and the applied cryptography, a pipeline which counts if the agent executives are to reach the production scale through the contact points of consumers and the company.
Crossarket, the objective of the machine economy is not limited to a single battery. The avalanche hosts the liquidity given by Ai-Governed via Blackhole Dex, Ethereum focuses on identity and regulations, accommodation of applications on the nearby chain and ICP on the chain and low-latency inference, and make GPU resources for training and model service.
The coverage of the metrics of Koinly and tokens place them in complementary roles rather than direct substitutes, by a thesis according to which the request for decentralized inference and the coordination of the market extends as the agents become of default actors in payments, realization and protocol operations.
If the ICP growth model for native IA accommodation is maintained, the chain inference cycles could reduce the latency by half by 2026, which would make the interactivity of viable agents for user -oriented applications such as intention routers, real -time coverage and the supply chain or IoT.
| Protocol | Primary AI function | Chain or TVL volume, September 2025 | TERM FOCUS |
|---|---|---|---|
| Ethereum | Announcement and regulation of the agent, ERC-8004, DAI team | $ 38 billion + | Confidence and coordination layer for agents |
| Bittensor, Tao | Decentralized training and inference markets | $ 1.4 billion is. | Open calculation exchange |
| Fetch.ai, fet | Autonomous economic agents, DAPP infrastructure | $ 640 million. | Machine machine coordination |
| RNE, RNDR | Decentralized GPU and inference | ~ 985 M $ | Calculate the spine for ai on a chain |
| Internet, ICP | Accommodation of the AI Native chain application | $ 800 million + | Lower latency for agental DAPs |
| Blackhole Dex, avalanche | MA and golden liquidity of AI | $ 193 million | Agent trading without permission |
The scenarios fall into three buckets.
A basic case in Ethereum consolidating the identity and confidence layer, because at least a quarter of the new DAPPs adopt the automation of agents by 2026, converging governance, the treasures, fees and payments into programmable policies anchored in the assignments.
A bull path turns on a more complete machine saving where the agents manage bilateral negotiation and the realization in the contexts of consumers and businesses, TVL DEFI going beyond $ 300 billion and decentralized API markets reaching a critical mass for long-tailed services.
A bear case focuses on the regulatory licenses of agents and the continuous centralization of calculation and access to the model, which would cap open participation and strangulation innovation in a small number of well -funded teams.
The overview of DLA Piper and Politics trackers indicate that the standards of transparency and control as a point of support, and not of pure and simple prohibitions, but the centralization of the calculation remains a known constraint.
Investors and manufacturers pass from tokens stories to measurable adoption triggers.
As for standards, the ERC-8004 is a basic surveillance element, as portfolios and childcare providers will have to implement certificate controls, recovery flows and the application of policies so that agents operate in complete safety in consumption contexts.
On the payments side, the AP2 pilots, if they were extended in large -scale cryptographic rails, would provide the first reproducible model for subscriptions, the invoicing of use and the realization between non -human actors, and would put pressure on bridges and accounts of abstraction of accounts to expose the limits and approvals with fine grain.
On the security side, the evidence in the field that adaptive controls reduce the losses made would unlock more autonomous governance, in particular for the setting of parameters on the volatile markets. Each of these tracks has public milestones which can be monitored without relying only on price tables.
The open question goes beyond knowing if the agents will transform; This is where the regulations and trusted checks occur.
If identity, certificate and policies live on the chain, the economy of the machine will be by default public registers, and DEFI will become the operating system for non -human economic activity. If these checks remain in closed platforms, the role of crypto will collapse in bridges and payment rails.
With the mandate of the DAI of Ethereum, the AP2 way for agents’ payments and a measurable change in the hiring of developers to crypto roles of AI X, the center of gravity is oriented towards the verifiable chain coordination which treats agents as first -class participants on the markets.
