Stablecoin-powered financial platform Fasset says it plans to become a licensed bank within three years, and that its banking services will extend not just to human customers but to the AI agents acting on their behalf. The company, valued at $1 billion after an August Series C, made the pitch in an interview its chief executive gave to Investing.com on October 2, 2026.
The framing is unusual among stablecoin issuers. Most competitors treat crypto rails as a way to move money faster. Fasset argues the real opportunity sits one layer up, in the intelligence that decides when and how money moves, and that the growth of autonomous software agents will force financial infrastructure to serve non-human customers directly.
“The system we are building has to work for that human, but increasingly for an agent acting on that human’s behalf, with intelligence embedded into how money, liquidity, and assets actually move,” the chief executive said, calling that intelligence layer the part most people underestimate.
What Fasset actually runs
Fasset operates Own Network, a regulated financial network that connects banks, telecom companies, payment providers, liquidity providers and custody partners across roughly 100 banking corridors in 125 countries. The company reports more than $40 billion in annualized transaction volume, over 3 million wallets, and more than 1,000 enterprise clients. Its Series C round raised $68 million led by Japan’s SBI Group, bringing 2026 fundraising to $119 million after a $51 million Series B earlier in the year that added Speedinvest to the cap table.
The company calls itself an AI-powered stablecoin neobank, a label that covers several distinct functions. Stablecoin settlement moves value across borders without correspondent banking chains. Tokenized asset infrastructure lets clients hold and transfer fractional claims on real-world assets. Corridor banking connects local payment systems in emerging markets to international liquidity. The newest layer, agentic AI, is where the company says its future competitive edge sits.
Why AI agents need their own banking
The argument for agent-facing finance runs something like this. As more commerce happens through software that researches, negotiates and executes transactions on a person’s behalf, the current account model, built around one human identity with one login, starts to break down. An agent buying flights, paying subscriptions, or settling invoices needs its own spending authority, its own limits, and its own audit trail, distinct from its owner’s card or wallet.
Banks have started to notice. Fasset is not the first to flag the problem, but few stablecoin platforms have made agentic payments a central product thesis rather than a feature. Circle’s Arc blockchain, for instance, launched a mechanism earlier this year letting AI agents pay human workers in USDC, an early test of machine-to-human settlement. Fasset’s version is broader: rather than a single payment rail, it wants the whole account structure, from identity to credit limits to compliance checks, to accommodate agents as first-class customers.
The regulatory path is the hard part. Banking licences in most jurisdictions require know-your-customer rules written around natural persons, and agent identity, agent liability, and agent authorization limits are all still unsettled questions. Fasset has not said which jurisdiction it will pursue the licence in or whether existing digital-asset banking charters would cover agent-run accounts.
The stablecoin context
Fasset’s pitch lands in a market that has split between two visions of stablecoin finance. One treats stablecoins as settlement infrastructure, a faster version of SWIFT that banks and fintechs plug into. The other treats them as consumer banking replacements, apps that hold deposits, make payments and offer yield without a traditional bank in the loop. Tether and Circle dominate the first model through sheer scale, with USDT still processing more than $70 billion in daily volume. Neobanks like Fasset and a wave of smaller competitors are betting on the second.
Both approaches are converging on the same regulatory pressure. The Fed proposed a two-business-day redemption ceiling for stablecoin issuers in late September, a rule that would force issuers to guarantee conversion to dollars within that window and effectively requires reserve portfolios heavy on short-term Treasury bills. The EU is weighing whether bank deposit rules should apply to stablecoin products that mimic deposit accounts, and Circle has lobbied Brussels to soften those requirements rather than face bank-style capital charges.
Fasset’s banking-licence goal reads as a bet that the consumer-style model will eventually be regulated like banking, and that the firms who get there first with licences and infrastructure will keep the lead once the rules settle. Competing without a licence in that scenario becomes difficult, since consumer demand tends to concentrate around the providers that can legally call themselves banks.
What would have to go right
The three-year timeline is aggressive. Banking charters typically take longer than three years to obtain, even in jurisdictions with fast-track digital-asset regimes. Wyoming’s SPDI charter, one of the fastest US paths for digital-asset companies, still took applicants 12 to 18 months through approval, and full-purpose banking licences in major markets routinely run longer. Fasset’s existing licences, described in its August funding announcement as covering emerging-market corridors, do not automatically convert into full banking status in major markets.
The agentic payments thesis also depends on adoption that has not arrived yet. Very few consumers currently let software agents initiate payments on their behalf, and the security model, what happens when an agent is compromised or makes a bad decision, remains an open question across the industry. Fraud liability, dispute resolution and agent authorization revocation all lack standard practice today. standards built around natural persons offer no clear answer to an agent spending $5,000 on a flight the owner never asked for.
There is also a competitive question. Major banks, card networks and payment processors all have AI agent initiatives in various stages of development, and none have committed to agent-first accounts as a flagship product. Whether a stablecoin neobank moves faster than the incumbents, or whether the incumbents simply absorb the feature once agent payments become standard, is the real risk Fasset is running.
Still, the company has something most stablecoin startups lack: real volume across 125 countries, real licences in multiple emerging-market corridors, and a strategic backer in SBI Group, one of Japan’s largest financial conglomerates. Whether the agentic banking thesis proves out or stays a pitch deck, the company is positioning ahead of a shift most incumbents have not yet built for, and the next three years will show whether that bet pays.