Ritual, an artificial intelligence research company founded by former Polychain Capital partners, has launched Ritual Chain, an Ethereum Virtual Machine (EVM)-compatible Layer 1 blockchain designed to support autonomous AI agents. EVM compatibility means smart contracts written for Ethereum can run on the network, allowing developers to reuse familiar tools and existing code rather than build from scratch. The network introduces native AI tools, confidential computing, and cross-chain transfers as demand grows for blockchain infrastructure that can host artificial intelligence applications operating largely without human supervision.

Built for Persistent, Independent Agents

Ritual Chain enables AI-powered applications to access models, process private data, manage funds, and schedule tasks directly through blockchain infrastructure, reducing the need for continuous human oversight. The launch reflects a broader effort to bring artificial intelligence capabilities into blockchain execution environments, allowing developers to build applications that perform complex operations independently.

Unlike conventional applications that typically depend on ongoing user input or centralized services, agents operating on Ritual Chain are designed to maintain their state and continue executing tasks over extended periods. The architecture is intended to accommodate AI agents that can function for months or years. Each agent can maintain its operational state, control assets through a non-custodial wallet, and continue working even when its creator is offline. The non-custodial structure places asset control with the agent itself rather than an intermediary, which is what allows it to transact without a person approving each step.

This approach could support applications requiring persistent execution, including automated financial operations, data processing, and other services that involve multiple steps.

Native AI Functions and Confidential Computing

The network introduces 15 native precompiles that allow smart contracts to interact with AI models and access capabilities that would otherwise require separate infrastructure. These functions include language model access, web data retrieval, private computation, and the scheduling of future tasks. Precompiles provide predefined operations that smart contracts can call without implementing every function from scratch, making it easier for developers to build applications that combine on-chain asset management with automated reasoning and task execution.

Ritual Chain also uses a distributed execution model in which specialized nodes perform computationally intensive operations while the blockchain verifies their results. This design separates complex computing workloads from the network's core verification process, potentially allowing applications to use advanced AI capabilities without requiring every blockchain participant to run the same calculations.

Privacy is another component of the architecture. Trusted execution environments are designed to protect sensitive information while it is processed by computing providers. These environments isolate workloads and help prevent unauthorized access to data during computation — an important consideration for applications handling confidential business information or private user records.

Infrastructure Partners and Cross-Chain Connectivity

Ritual Chain launched with support from several infrastructure and research organizations, including Phala Network, LayerZero, Stargate, and Nous Research. Phala Network provides confidential computing infrastructure, LayerZero supplies cross-chain messaging capabilities, Stargate supports transfers of digital assets between blockchain networks, and Nous Research contributes expertise in artificial intelligence research.

The mainnet went live with an announcement on X:

Ritual mainnet is now live. AI works inside your companies. Ritual lets AI be one.

— Ritual (@ritualnet) October 26, 2026

The integration with Stargate enables USDC transfers between Ritual Chain and other supported networks. USDC is a U.S. dollar-pegged stablecoin commonly used for on-chain payments, which makes it a practical unit for funding agents that need to pay for services or execute transactions. This connectivity could help users and developers move stablecoins across blockchain ecosystems, making it easier to fund AI agents and support applications that require access to assets held on different networks.

Cross-chain interoperability is particularly relevant to autonomous agents that may need to interact with decentralized finance applications, payment systems, or other blockchain-based services. The availability of stablecoin transfers provides a potential foundation for agents that execute transactions as part of their assigned tasks.

Testnet Activity and Security Review

Ritual said its public testnet has been operating since June and has processed more than 40 million transactions. The network also reports block times below 300 milliseconds, indicating an emphasis on rapid transaction processing. These figures provide an early indication of testnet activity, although they do not independently establish how the network will perform under sustained production workloads or at larger scale.

Blockchain security firm Zellic has audited the protocol, adding an external security review to the project's development process. Audits can help identify vulnerabilities in smart contracts and protocol implementations, but they cannot guarantee that a network is free of security risks.

Ritual Chain's next challenge will be translating its testnet activity and technical capabilities into reliable applications that developers and enterprises can use. Whether mainnet agents operate dependably outside testnet conditions, and how quickly developers put the network's native AI functions to use, will provide early evidence on that front. Its long-term adoption will depend on execution reliability, privacy protections, security, and the practical value of integrating AI agents with blockchain-based financial and computing infrastructure.