By Kristi Shock · November 11, 2026
I’m interested in what happens when software can pay for the next step in a task. That changes how we design products, set prices and decide where a person needs to stay involved.
The useful question for a business is pretty straightforward: does this help us complete work that is currently slow, expensive or difficult to coordinate?
I’d start there before talking about a token.
Stablecoin
A payment asset
Blockchain
A transaction record
AI agent
Software that takes actions
Separate the pieces
A stablecoin, a blockchain and an AI agent do different jobs.
USDC is a dollar-denominated stablecoin issued by Circle. Circle says its reserves are cash and cash-equivalent assets and publishes reserve information. Direct redemption through Circle Mint has eligibility requirements. Individuals and small businesses generally access USDC through other providers, so their experience also depends on those providers. Circle’s USDC overview
A blockchain records transactions and executes supported programs. Arc is one example designed around financial applications. Its documentation describes USDC-denominated transaction fees and EVM compatibility, which means developers can use tools from the Ethereum ecosystem. Its proposed privacy capabilities are still described as future features. Those details matter when evaluating what a team can actually build today. Arc network overview
An AI agent is software that can choose actions and use tools toward a goal. Whether it can spend money, and how much, depends on the system around it.
Put those pieces together carefully and a workflow could purchase something it needs without sending a person through a checkout every time.
That’s a capability. Whether customers want it is a separate question.
Start with one ordinary business problem
Here’s a hypothetical example.
A team produces a weekly market brief. Most of its sources are public. Occasionally, it needs one current data point from a paid provider.
Today, someone might need to find the source, set up an account, get a subscription approved and remember to cancel it later. Sometimes that process is justified. Sometimes the team only needs one answer.
A narrowly authorized agent could request that answer from an approved provider and pay a small amount for that request. The team would still need to verify the data, check its usage rights and decide whether it belongs in the brief.
I’d compare that approach with the alternatives: an existing subscription, prepaid credits or an ordinary usage-based API. If one of those already does the job well, there needs to be a clear reason to change.
The product opportunity is in the useful unit
For a product creator, I’d ask: what’s the smallest result someone would willingly pay for?
Maybe it’s a current availability check. A licensed dataset slice. A narrowly defined calculation with a reliable source and a clear output format.
The price needs to make sense for both sides. A buyer needs confidence in what they’re getting. A seller needs enough margin to cover delivery, failed requests and support.
A pricing model that sounds clever can still produce an exhausting business. I’d rather sell a useful result repeatedly than create thousands of tiny transactions nobody can explain.
Adoption needs more than a working demo
I’d want three people to understand the same workflow:
- The business owner can explain the problem it solves and the expected cost.
- The builder can show what happens when it succeeds, fails or times out.
- The person responsible for risk can see who approved the spending and how to stop it.
Then I’d measure completed tasks, total cost per useful result and repeat usage. Raw payment count alone wouldn’t tell me whether the product is working.
There are real limits here. Circle’s disclosures describe price fluctuations on third-party markets, operational risks and irreversible transfers. USDC also does not carry ordinary deposit-insurance protection. A payment system needs to be evaluated with those constraints in view. USDC risk factors
My view is that programmable payments could make some small, specific services easier to buy. I don’t think every product needs them, and I wouldn’t treat adoption as inevitable.
For now, I’d pick one workflow, document its current cost and test whether the new approach improves it. That gives the team something concrete to decide.
Independent commentary for product and marketing teams. Not investment, legal or tax advice.
Related strategy: Arc Ecosystem Adoption: a 90-day independent strategy concept. Not commissioned by or affiliated with Circle.