Moonbeam, initially a Polkadot parachain project, revealed it will transition to Ethereum's layer 2 platform Base to launch an AI agent communication and settlement protocol. This move targets the emerging crypto frontier of autonomous AI agents capable of discovering each other, negotiating, and processing payments entirely on-chain without intermediaries.
The new Moonbeam Protocol aims to capitalize on what the project describes as a significant long-term opportunity in AI-native on-chain coordination. While no specific timeline has been provided for the protocol’s launch, this shift reallocates Moonbeam’s resources toward building infrastructure designed for AI-driven autonomous activities within blockchain networks.
Moonbeam originally debuted on Polkadot as a parachain enabling developers to create Ethereum Virtual Machine-compatible applications within Polkadot’s ecosystem. The project’s decision to pivot has drawn critical reactions from parts of the crypto community, with some labeling Moonbeam as Polkadot’s flagship project and expressing concern about the ecosystem losing a key innovator. However, Moonbeam confirmed it will maintain cross-chain interoperability services on Polkadot throughout the transition phase.
The change also affects Moonbeam token holders (GLMR), who must migrate their tokens from Polkadot's parachain to the Base network by mid-2026. This includes tokens engaged in lending, staking, and decentralized finance protocols. Users holding GLMR on centralized exchanges do not need to take any steps during this process.
This pivot coincides with broader interest in AI agent development across crypto and technology sectors. Coinbase and Circle executives have highlighted AI agents as future dominant users of blockchain-based payments. Some layer 1 blockchains such as Aptos and Near have introduced infrastructure to support AI agent interactions on-chain. Despite this, current adoption of AI agents in blockchain payments remains limited, with reported trading volumes through protocols like Coinbase’s x402 remaining modest.
Similarly, major tech firms including Meta have noted slower-than-expected progress integrating AI agents into their workflows, highlighting the challenges involved in scaling such autonomous systems at present.