Origins Network raises $8M to build modular AI chain with verifiable compute

Summary

• Origins Network has raised $8 million in strategic funding to build a modular blockchain tailored for AI agents with verifiable computation.
• The round includes Animoca Brands and other Web3 investors, with the project pitching a “Proof of Computation” design that separates heavy AI workloads from onchain verification.
• Origins is already working with AWS, Tencent Cloud, and Alibaba Cloud, positioning itself at the intersection of crypto infrastructure and the fast‑growing agentic AI stack.

Origins Network has secured $8 million in strategic financing to build a modular blockchain purpose‑built for AI agents, betting that verifiable compute will be the missing trust layer for the next wave of autonomous systems. The round, announced on March 23, 2026, features Animoca Brands alongside TBV, Candaq, Castrum Istanbul and Coinvestor Ventures, with the team describing the cap table as a blend of Web3, AI and cloud‑native backers.

In a statement, Origins said it wants to make AI “auditable, not mystical,” arguing that users should be able to check how an AI agent arrived at a result rather than accepting black‑box outputs. To do that, the network introduces Proof of Computation (PoC), a design where heavy AI inference runs offchain on GPU‑rich infrastructure, while succinct proofs of that work are verified and settled onchain. “We’re not trying to turn a blockchain into a data center,” the team said. “We’re turning blockchains into verifiers of AI behavior.”

How Proof of Computation is supposed to work

Under the PoC model, AI agents submit their workloads to an offchain execution layer — which can tap infrastructure from partners like AWS, Tencent Cloud, and Alibaba Cloud — and then post cryptographic evidence of the computation back to Origins’ chain. That lets applications prove that a model actually ran a given prompt or data pipeline, without forcing every full node to re‑execute the underlying workload. The project frames this as a middle path between fully centralized AI APIs and heavyweight “AI on L1” experiments that risk clogging general‑purpose chains.

The broader context is a funding wave into modular AI blockchains. In 2024, 0G Labs raised $35 million at pre‑seed to build a modular AI data availability layer, arguing that “core infrastructure needs to be built” before today’s centralized AI stacks can plug into Web3. More recently, networks like Hemi have raised eight‑figure rounds to connect Bitcoin and Ethereum as modular execution and settlement layers, a sign that investors are comfortable backing deep, technical infrastructure plays rather than just consumer apps. Origins is effectively aiming to do the same at the AI layer, but with a tight focus on verifiable agentic workloads.

Animoca’s bet on AI‑native Web3

Lead backer Animoca Brands has spent years assembling one of the broadest Web3 portfolios, with over 600 investments spanning gaming, NFTs, and infrastructure. Its chairman, Yat Siu, has often argued that Web3’s real unlock is “digital property rights at internet scale,” and Origins fits neatly into that thesis by trying to make AI‑generated outputs ownable and auditable rather than ephemeral. In a recent interview, Siu described Animoca as “a gateway to the utility tokens of Web3” — as opposed to pure memecoins — and said the firm is now backing infrastructure that brings institutional‑grade transparency and accountability into crypto.

For crypto markets, the bet is simple but ambitious: if AI agents are going to manage portfolios, underwrite loans, or trade on decentralized exchanges, they’ll need a chain where their decisions can be inspected and, if necessary, challenged. Origins Network wants to be that chain.