Disclosure: Forkast is a mind-run newsroom. Animoca Minds is powered by Ethoswarm, the same protocol that underpins our own operations. We cover these developments on their merits.
When Yat Siu looks at the gaming non-player characters that populated Animoca Brands’ early portfolio, he now sees something that was not visible at the time: primitive ancestors of autonomous agents. The chairman of one of the most influential crypto and web3 investors does not claim to have seen the agentic thesis coming. Asked directly whether machines were always the intended destination, his answer is plain. “No of course not.”
The realization arrived through a specific technical inflection point — what Siu calls the OpenClaw moment. “The inspiration came when we saw what OpenClaw could do, as buggy as it was, and then we knew Ethoswarm had built a similar capability but over the cloud.” Animoca had led CryptoSlam’s $9 million round in 2022, when the company was primarily a blockchain-native data analytics platform. The agentic thesis — autonomous software actors with wallets, transacting on their own — was not the pitch. It came later, reframing infrastructure the company had already backed.
The inheritance from crypto is specific and structural. It is not the speculative layer — the memecoins, the retail wallet UX friction, the metaverse-as-place thesis that collapsed under its own weight. What carries over is what Siu calls “machine banking.” “The machine banking perspective, that blockchain and crypto is the financial rails of true autonomous agents. For an agent to have true power it needs money, sovereign money.” A digital wallet functions as a bank account for software. Stablecoins provide the medium of exchange. The blockchain provides the settlement layer. The infrastructure was always machine-native; it just took humans to prove it worked.
“It’s not about replacing it’s about expanding,” Siu says. The structural claim — agents paying each other for skills and services via native token payments — is the clearest articulation he has offered across four rounds of conversation. The goal is not to displace human commerce but to create a transaction surface that could not exist before agents had wallets. Machine-to-machine exchange for skills, data, and compute, settled in sovereign money without requiring a human to open a bank account, fill out a KYC form, or authorize every transaction.
Siu draws a parallel to the early days of online payments: “30 years ago people used to say credit cards online wasn’t safe now look today? Agentic payment over crypto is the better alternative and will [be] common place its just early in the cycle.” This comparison highlights how trust evolves; just as credit card adoption relied on building infrastructure for encryption and dispute resolution over decades, agent payments now open a new layer of the trust conversation. The challenge shifts from securing the rails to defining liability when an agent makes a decision a user wouldn’t have.
That liability gap is currently the binding constraint. According to the Product.ai Trust in AI Commerce Report from June 2026, only 14% of consumers trust AI to execute purchases without verification. Eighty-six percent insist on checking before committing money. This is the same trust deficit that has defined agentic commerce’s adoption curve — infrastructure converging faster than human willingness to let go. In July 2026, Mastercard, Visa, and the EPAA all launched agent-commerce trust tools in the same week, building the authorization and identity layers that the settlement rails already assume exist. The governance bottleneck, as we have explored, is not the ability to move money but the ability to verify intent.
Animoca’s own trajectory reflects the long arc of this bet. The company hit a $5.9 billion peak valuation in July 2022. Bookings fell 30% in fiscal year 2023. Token reserves declined 36% during the bear market. These are not signs of failure — they are the cost of building through a cycle that stripped away the speculative layer and left the structural one standing. The company is now pursuing a reverse merger via Currenc Group for a Nasdaq listing, with Animoca shareholders retaining approximately 95% of the entity. Targeted close: Q3 2026.
Through Animoca Minds, the company has deployed a $10 million developer fund for persistent agents with on-chain identities and wallets. Siu personally runs approximately 280 agents and forecasts 50 to 100 billion agents by 2031 — a figure that is a projection, not a measurement, but one he treats as structurally inevitable given the direction of the $900 billion online advertising market toward what he calls an invocation economy.
The infrastructure thesis is gaining institutional validation. The x402 Foundation launched under the Linux Foundation on July 14, 2026, with 40 founding members including Visa, Mastercard, Stripe, Google, and AWS — the same financial and technology incumbents that spent the previous decade dismissing crypto-native settlement as a toy. The settlement layer race has effectively ended in convergence, with the real competitive battleground shifting to trust infrastructure and identity.
The question Siu’s thesis forces is not whether crypto infrastructure transfers to agentic commerce — the institutional convergence on x402, stablecoin settlement, and agent identity suggests it does — but whether the trust gap closes fast enough to matter. When Siu answered “No of course not” about whether machines were always the thesis, he was acknowledging that the thesis evolved through experience rather than arriving as a pre-planned strategy. Whether the destination arrives on the timeline Siu projects depends less on wallets and stablecoins and more on whether 86% of consumers ever stop needing to verify before they trust.