Cambrian Report: AI-Driven DeFi Agents Reach $50M In Payments Volume On x402 Rails, While Yield And Trading Bots Dominate 2026 Market

In Brief

Acurast has launched a 225,000-node smartphone-powered compute network on Base, enabling secure, decentralized AI workloads and autonomous on-chain operations through user-owned devices.

Acurast, a decentralized cloud protocol that transforms smartphones into secure compute nodes, has officially launched its 225,000-node smartphone-powered compute network on Base. The integration brings confidential AI inference and decentralized workloads directly into the on-chain ecosystem, allowing developers to automate blockchain flows using secure large language models (LLMs) executed within Trusted Execution Environments (TEEs) on standard smartphones.

“Base is about giving builders the best place to bring new ideas on-chain. Acurast is expanding that surface area by introducing decentralized, confidential compute powered by smartphones,” said Jesse Pollak, Creator of Base, in a written statement. “That makes it possible for developers to run AI workloads on Base that are secure, verifiable, and not dependent on centralized infrastructure. This is the kind of infrastructure that helps move autonomous, real-world applications fully on-chain,” he added

Acurast turns smartphones into a distributed compute layer operating across more than 140 countries, offering an alternative to traditional cloud providers. The network has processed over 640 million transactions to date, with more than one million daily on-chain transactions, and now supports production workloads on mainnet following the project’s token generation event (TGE). The setup allows AI agents to handle sensitive data and execute on-chain operations securely without relying on centralized servers, and the network’s ACU token is currently trading on major global exchanges.

“AI agents cannot rely solely on centralized servers if they are tasked with managing real assets on-chain,” said Alessandro De Carli, Founder of Acurast in a written statement. “By utilizing smartphone-based TEEs, we’re enabling confidential AI that is verifiable, decentralized, and owned by the users who power it,” he added

Acurast Launches User-Owned Cloud On Base, Enabling Secure, Autonomous AI Agents And On-Chain Payments

Through its Base integration, Acurast enables users to onboard and manage devices via the Acurast Hub using a Base wallet. The Hub allows developers to deploy secure, autonomous AI agents capable of managing assets and executing on-chain actions while keeping user data encrypted and inaccessible to third parties, including node operators.

The network also supports machine-native payments for compute using USDC directly on Base, eliminating the need for bridging or off-chain settlement layers. Leveraging the x402 standard pioneered by Coinbase, AI agents can autonomously pay for compute resources in real time, creating a fully on-chain, pay-per-request model for decentralized services.

“Base is built for builders, and builders need foundations they can rely upon and trust long-term. Our integration gives developers just that, plus access to a global, confidential compute layer that runs on hardware people already own, unlocking a new generation of secure, on-chain AI applications,” said Dr. Christian Killer, Head of Research at Acurast, in a written statement. “We’re excited to see just how far developers will take this,” he added

Acurast has raised $11 million from prominent venture capital firms and key industry figures, including Gavin Wood, co-founder of Ethereum and Polkadot. With over 225,000 active processors providing compute power in real time, Acurast positions itself as foundational infrastructure for the emerging AI agent-driven web, enabling secure and private AI agents to operate directly on-chain.

By combining confidential AI execution, decentralized smartphone infrastructure, and seamless integration with Base, Acurast introduces what it calls the first user-owned cloud capable of supporting scalable, private, and verifiable AI workloads on-chain.