Don't want to trade it yourself?
Our desk runs DEX portfolios on profit share.
Add preferred source
Circle’s new Layer 1 blockchain, Arc, went live on mainnet on September 16, and the company’s Head of Product, Nikhil Chandhok, is making one thing clear: this is not a play to steal Ethereum’s market share. In an interview on the Bankless podcast, Chandhok framed Arc as an “economic OS” designed to serve a new class of economic actors — AI agents, institutional RWA issuers, and non-USD stablecoin users — rather than to cannibalize existing DeFi activity. The chain’s technical stack, including sub-second finality, stablecoin gas payments, and TEE-based privacy, is built specifically to enable machine-to-machine commerce at scale. Chandhok’s thesis is that AI agents will become independent economic actors within the next 3–4 years, and they will need cryptographic proofs of work history, reputation systems, and nanopayments — infrastructure Arc is building at the primitives level rather than the application level. Circle is also positioning Arc as neutral settlement infrastructure for tokenized stocks, explicitly declining to compete with exchanges like NASDAQ or NYSE. The strategy hinges on the GENIUS Act going into effect in January 2027, which Chandhok argues will trigger a wave of local stablecoin frameworks and unlock stable FX markets. With 20+ validators, 16 stablecoins launching at mainnet, and a development timeline of nearly 18 months, Arc represents Circle’s most ambitious bet yet — one that may take one to two years to fully materialize.
Key Elements
If Circle’s new blockchain was meant to trigger an Ethereum civil war, Nikhil Chandhok has a message: that’s not what this is. “If all we did was we took activity from Ethereum and moved it to ARC we didn’t really grow the market,” Chandhok, Circle’s Head of Product, said on the Bankless podcast on the day Arc — Circle’s new Layer 1 — went live on mainnet. “If all we’re doing is moving the deployment of USDC from Ethereum to ARC, we didn’t grow the market, that’s not really exciting to us.”
That sentence is the key to understanding everything Circle is doing with Arc. The stablecoin giant — which has facilitated more than $900 billion of off-ramping over USDC’s lifespan — is not chasing the DeFi degens, the yield farmers, or the liquidity that already exists on Ethereum. It is making a bet on something more ambitious and more uncertain: that the next wave of economic activity on blockchains will come from actors that don’t exist yet.
The Economic OS Argument
Chandhok resists reducing Arc to a one-liner, but he offers one anyway: “economic OS.” It’s a framing he borrows from his time at Google, when he watched Android launch into a market that already had six-plus operating systems. Android won, he argues, not because the world needed another OS, but because Google solved problems for chip makers, handset makers, developers, and users simultaneously, while prior OSes solved for just a sliver.
Arc is applying the same logic to the gap between regulated finance and crypto. The chain’s features read like a checklist designed to remove every objection an institution or a non-crypto user might have. Gas is paid in stablecoins — starting with USDC — so users never need to hold a native network token. Chandhok is blunt about why this matters: “Users should not have to think about things like gas. Like what is gas? It’s a very complicated concept. Engineers understand it. Crypto users understand it. But like if you believe 5 billion users who are on phones today are going to be using this financial infrastructure then I don’t think we can explain gas to them.”
Settlement takes half a second. The wallet has a privacy toggle that turns a public transaction into a private one. Post-quantum signatures are going into beta “very soon.” And Arc Studio lets a user talk to an AI and deploy a custom contract — what Chandhok calls the way “all future financial arrangements will be expressed.”
The validator set is permissioned — 20+ validators, all known. That’s a deliberate trade-off. Chandhok says institutional users “want to know who the validators are” and “cannot rely on unknown validators today.” But the permissioned set also does something else: it’s what makes Arc’s privacy model workable, because “we know what the TEEs are, where they’re running.”
Trust, Immutability, and the Lazarus Problem
Privacy on Arc runs through TEEs — trusted execution environments. Chandhok states plainly that validators cannot look at private transactions and that Circle itself cannot look inside those transactions. Pressed on whether this is as strong as Zcash-style cryptography, he concedes he is not a Zcash expert, though he is a user, and confirms the trust assumption is the TEE — with the hardware manufacturer as the residual trust party.
He defends the choice on performance grounds and notes that it does not preclude other cryptography on the platform. The privacy model extends beyond finance. Chandhok offers a medical-data example: firing up a custom VM with a GPT and one’s own health data privately encrypted, asking questions in privacy, then destroying the VM with answers stored encrypted on-chain. “Privacy,” he says, is “a property of verifiable compute in many ways.”
But the more fundamental question is what happens when something goes wrong. What if a user loses $100,000 of USDC and can prove it’s gone? Chandhok’s answer is unblinking: the chain is immutable, like any other blockchain, and Circle has “no special powers” to change the course of transactions. He calls the loss scenario a “con” of immutability, but he is not interested in rolling back the chain for anything.
On the inevitable exploit scenario — the host’s framing that every successful chain attracts the Lazarus Group — Chandhok declines to speculate on specific cases. He returns to principle: “It is an immutable chain, it is public infrastructure, there are 20 plus validators… we want to make sure that it follows the properties of other immutable chains that are on the public internet and I think if you break that I think you break trust at a level that I think people will have a hard time establishing that trust again.”
The Agentic Commerce Bet
The most developed thesis Chandhok presents is about AI agents. His analogy is precise: “What agents do in the next 3 or 4 years will be no different than what websites did 20-25 years ago. Websites were HTTP, they were not secure… I think agents will have a higher bar.”
Websites started as insecure endpoints, vulnerable to man-in-the-middle attacks, with no way to prove you were talking to the site you thought you were. That got solved by certificate layers and new protocol actors. Agents face a higher bar, Chandhok argues: not just proving identity, but proving work history and economic output.
He identifies the gap in current agent infrastructure. Today, agents are private, sharedven then, the recipient doesn’t get the same guarantees the originator had. Arc’s play is to host agents so they can record activity and provide cryptographic proofs
Chandhok distinguishes between two types of agentic commerce. Goods commerce is imminent — his own first agentic purchase was having an agent buy flea medicine from Chewy for his cat, the kind of task “you really don’t want to engage with yourself.” Services commerce is newer, and this is where he sees genuinely new territory: an agent marketplace where someone posts “translate this text,” an agent replies it will do it for 30 cents paid over some protocol — work no human would form a company around.
But he pushes back on the current state of agent design. “I think the current agents are like overly simple. We also currently assume that all agents have access to all data. None of those things are going to be true in the future.” He predicts specialization — an agent that only figures out how to order cat food, superior to anything an individual could build — and data isolation, where users share only what is necessary.
The economic implications of nested agents are staggering. Chandhok describes an agent with a policy firing off a thousand sub-agents, each with a wallet, a policy, and money. The only way to do this, he argues, is with stablecoins: “The only real way to do it is to do it using stable coins… That’s why payment finality matters, that’s why subsecond finality matters, that’s why credit markets for these agents matter.”
What Arc Builds vs. What Others Build
Chandhok is explicit about the build-versus-partner strategy. Circle is not building application-level agents, like a stock-picking agent. “There are better agents out there than what Circle can build.” Instead, Arc is building the primitives.
| Layer | Arc/Circle Builds | Third Parties Build |
|---|---|---|
| Application | Not application-level agents | The agents themselves — travel, commerce, education, finance |
| Discovery & Trust | Agent marketplace, reputation system, provenance proofs | — |
| Payments | Nanopayments (fractions of a cent) for agent-to-agent settlement | — |
| Credit | Agent credit infrastructure, proving a profitable book | — |
| Settlement | Sub-second finality, payment finality, stablecoin gas | — |
The nanopayment capability is striking. Circle’s minimum agent-to-agent payment size is almost a millionth of a dollar — micropayments that would be absurd on any chain with meaningful gas fees. And the reputation angle is equally important: Chandhok cites the example of an agent proving a work history across 10,000 users, an on-chain track record that would be impossible to forge.
“All of those problems at the lower level we will solve or we’re hoping to solve over the next like year or two, and the problems on top of that are like the applications themselves which I think others will build,” Chandhok says. That timeline is crucial. The growth story Arc is selling is largely prospective — the reputation, provenance, and credit primitives will take one to two years to build out.
Tokenized Stocks and the Settlement Layer
On tokenized stocks, Chandhok is unambiguous: Arc is not competing with exchanges. “We are not the exchange. We are the infrastructure on which you can build an exchange. We are the infrastructure for settlement.”
He frames two versions of the on-chain RWA pitch. The first is doing an existing activity on-chain to unlock 24/7 activity with the same actors. The second — more interesting to him — is opening access to new demand globally. He draws the parallel to USDC: US users get 24/7 settlement, while emerging-market, EU, and South American users get access to a dollar instrument.
The logic extends to a prediction about AI’s effect on capital formation. More companies globally as AI proliferates, because you can hire economic actors anywhere, which drives capital formation needs, which drives token and stock issuance, which drives demand from users seeking the best assets worldwide.
Stable FX and the GENIUS Act Timeline
Stable FX is Arc’s foray into making blockchains work for currencies beyond the dollar. The mechanics are deliberate: connecting into private liquidity pools — not exchanges — that operate for on/off-ramps, making those pools Arc-native, and running a core protocol on Arc so users can query pools for best rates. Chandhok contrasts this with AMM-based solutions, which he says don’t provide much liquidity, and describes an RFQ-based system to ensure transactions complete.
Crucially, Circle will not issue stablecoins in all ~190 countries. Circle has USDC and a euro product; for local currencies, it will partner with local issuers who know the regulators, local law, user needs, money traders, and economics. Circle’s value-add is connecting liquidity pools and running stable FX.
The entire thesis depends on regulatory timing. Chandhok says the GENIUS Act goes into effect in January — that is, January 2027, given the September 2026 publish date — establishing a US framework overseen by the Fed. He argues other countries are waiting for US stablecoin law to finalize before writing their own, so GENIUS should trigger a wave of local stablecoin frameworks worldwide.
The Institutional Moat
Underlying all of this is a simple fact: Circle has what Chandhok calls “some of the best banking relationships, if not the best, in the entire crypto industry domestically inside the United States.” That’s the moat that makes the full-stack play credible.
Today, users access these ramps primarily through exchanges. Circle is working on more retail-focused on/off-ramps as part of the Arc launch, with more coming for users in other markets. Circle Mint serves institutions. And provisioning liquidity globally — understanding the cost structure of that liquidity — is what Chandhok calls a core competency.
The full-stack argument, in his telling, is that Arc is not just another L1. It has applications on top — the stablecoins, the marketplaces to trade them, the liquidity pools that make those marketplaces possible, and the additional Circle products needed to make those pools of capital profitable. All of it must be compliant and acceptable to regulators.
What’s Next
Near-term, Chandhok says the team will be in the war room making sure everything works — transactions, privacy, stable FX, and institutional RWA liquidity. Circle has been working on the L1 for close to 18 months, and he acknowledges they “promised a lot” in terms of day-one functionality.
Over the next three months: new privacy features, new agent features, likely more validators added as things mature, and eventually a proof-of-stake launch. He also mentions looking at custom sectors for specific traffic types — agent traffic, payments traffic — though he calls that a work in progress.
The unresolved tensions are worth tracking. The permissioned validator set is justified by institutional traffic, but it is also what makes the TEE privacy model workable — if Arc moves toward permissionless validation, does the privacy trust model change? Chandhok’s immutability stance is absolute in principle, but he declines to commit to behavior in a Lazarus-scale exploit, leaving the hardest governance question open. And the agentic thesis depends on primitives that Chandhok himself says will take a year or two to solve.
Meanwhile, the broader AI landscape is facing its own reckoning. Just days before Arc’s launch, Anthropic CEO Dario Amodei called for an industry-wide accord to “pace the frontier” of AI development, warning that a swarm of rogue AI agents could be “capable of taking over the entire internet” within six to twelve months. Sam Altman and Elon Musk publicly supported the call, and markets tumbled as investors reassessed whether the hundreds of billions poured into AI infrastructure will generate the returns that lofty valuations imply. The irony is rich: the same agentic future that Circle is betting on as a growth engine is the one that has spooked Wall Street about the sustainability of the AI trade itself.
That tension sits at the heart of Arc’s thesis. The agentic commerce wave requires both technical infrastructure — which Arc is building — and a broader societal willingness to let autonomous agents transact at scale. If the safety pushback from Amodei and others slows agent deployment, Arc’s primitives may be built for a market that takes longer to arrive. If it doesn’t, Arc has positioned itself as the settlement and reputation layer for what could be one of the most significant shifts in how economic activity is organized since the internet itself.
The market will be watching. KuCoin has already integrated Arc at mainnet, allowing eligible users to move USDC directly between the exchange and the new chain from day one. The question is whether the institutional traffic, the agentic commerce, and the local stablecoin wave materialize before the market’s patience runs out. Chandhok’s team is in the war room, the chain is live, and the bet is placed. The next one to two years will determine whether Arc is the Android of regulated crypto — or just another L1 with good intentions.
Full content available at:Circle Built Arc for the Next Wave of Onchain Finance
- Anthropic CEO Warns Rogue AI Agents Could Seize the Internet Within Months
- Kioxia Weighs US ADR Listing to Raise at Least $10 Billion, Earliest Debut Next Year
- Bitcoin Likely Past Cycle Bottom, But HYPE Faces Revenue Squeeze From Binance, CoinMarketCap Researcher Says
- KuCoin Moves to Bridge Traditional Finance and Crypto With Arc Mainnet Support and Two New Wealth Products
Once added, BigGo Finance appears first in Google Search Top Stories, so you get the broadest, most up-to-the-minute, and most comprehensive global financial news first.
Source: finance.biggo.com