Developers post live apps, integrations, or demos.

Ecosystem Showcase & Launches

Know Your Agent Was a Category in July. In September, It Got a Deadline.

A category tells you what to build. A date tells you when you are late.



On 29 September 2026, Mastercard published a report called A Short History of the Future of Shopping and Payments. Most of the coverage picked up the headline number: more than one in ten online shoppers routinely using AI agents to buy things by 2030.



That is the interesting number. It is not the important one.



The important one is buried in a list of five predictions written by Theodora Lau, one of four futurists Mastercard commissioned for the report. Prediction three reads: Know Your Agent becomes a compliance advantage. And then it puts a date and a counter on it.

By 2030, at least three G20 financial regulators will issue formal guidance on how AI agents must be registered, authorized, monitored, and held accountable. First movers that can build the KYA infrastructure will hold a compliance advantage.



Read that twice, because two separate things just happened in one sentence.



The first is that the phrase left the vendor pitch deck. Know Your Agent has been a product category since roughly July, when payment companies started launching registries and scoring products under that name. Categories are cheap. Anyone can name one.

The second is that somebody attached a clock to it. Not a trend, not a direction of travel. Three regulators. Four years. Registered, authorized, monitored, held accountable.






Be precise about the provenance, because it changes how much weight the sentence carries.

Mastercard published the report. Theodora Lau wrote that prediction. Those are different acts. Lau is an independent author, one of four futurists contributing alongside Magnus Lindkvist, Patrick Dixon and Katja Forbes. Mastercard's own executives wrote separate sections under their own names, including Greg Ulrich, the company's Chief AI and Data Officer, Pablo Fourez, Gaurang Shah and Sherri Haymond.

So this is not a Mastercard regulatory forecast, and it is certainly not a Mastercard requirement. 



What it is: a global card network chose to put that sentence in a document with its name on the cover. Networks are careful about what they publish. The choice to include a dated compliance prediction about agent identity, in a report about the next four years of commerce, is itself the signal.

And Mastercard's own contributors are pointing at the same wall from the other side. Greg Ulrich writes that as AI agents become more capable, a new question is emerging: not what they can do, but whether people will trust them to act on their behalf. Sherri Haymond writes that trust in agentic commerce requires knowing the agent behind a transaction, verifying intent, enabling permissions, and applying the intelligence to detect new forms of fraud.



Different authors. Same conclusion. The bottleneck is not capability.



And then, the day after the report, Mastercard stopped describing the problem and shipped against it.

On 30 September, at its RiskX security conference, the company announced an expansion of Agent Pay with new trust and intelligence services. Its Agent Pay Trust Framework is built on five components: identity, intent, controls, execution and intelligence. What shipped is the intelligence layer, and the first service in it, now in testing in the United States, returns a probability score that a transaction was initiated by an AI agent.

That is worth reading precisely. A score telling a merchant a transaction probably came from an agent is a detection signal, and a network shipping one in September rather than in 2030 is the strongest evidence available that Lau's timeline is not conservative. It is also not a mandate and not a receipt. Whether the agent is software is the first question. Whether it was authorized, and whether the thing it paid for arrived, are the two that decide who carries the loss.




A category is a marketing fact. You can enter it whenever you like, and you can leave it just as quietly.



A regulatory timeline is an engineering fact, because compliance infrastructure has a build time that does not compress. An identity registry that a supervisor will accept is not a weekend project. Neither is an audit trail that survives a supervisory review, nor a permission model that can prove, after the fact, who authorized what and when.

Firms that start building the month a rule lands do not ship on time. They ship late, and they ship the version they could finish rather than the version they wanted.

This is the part of Lau's prediction that deserves the attention. Not the forecast itself, which may be right or wrong. The observation attached to it: first movers hold the advantage. In compliance infrastructure, first mover advantage is not a land grab. It is simply having built the thing before the deadline existed.




The stack that answers it already exists, in public

Here is the part that should make this less abstract for anyone building on Arc.



The pieces this prediction describes are not hypothetical, and they are not proprietary to anyone. They are public standards, written in the open, with contributors from the Ethereum Foundation, Circle, Coinbase, Google and the card networks.

ERC-8004 answers who the agent is and what its record is. ERC-8183 answers what happens when a paid job does not complete. x402 answers how value moves between software at a size a card rail will not touch. x401 answers who authorized the agent in the first place.

Registered. Authorized. Monitored. Held accountable.

Those four words from the prediction map onto those four standards almost one to one. That is not a coincidence, and it is not clever positioning. It is what happens when two groups of serious people look at the same problem and independently arrive at the same list of unanswered questions.

We have written the full map of those four standards before and will not repeat it here. 
The point today is narrower: 
the infrastructure a global network expects regulators to ask for by 2030 is being specified, in public, right now, by people who publish their work.






None of it settles anywhere useful without a unit of account a finance function can hold, and a chain where the fee does not require a second asset.

Arc's mainnet opened on 16 September 2026. It is Circle's Layer 1, engineered for institutional finance, and its design choice that matters most to agent payments is the most boring one available: 
USDC is the native gas token. An agent settling a payment does not need to acquire and manage a volatile asset to pay the network fee. One asset in the account. One line in the books.

That sounds like a convenience. At agent scale it is a structural requirement. Software making thousands of small payments a day cannot run a treasury policy for a token it never wanted.

Arc launched with more than a hundred enterprise integrations and a validator consortium that reads like a list of people who do not join things lightly: BlackRock, Visa, Mastercard, DTCC, Standard Chartered, Intercontinental Exchange, MoneyGram, SBI Group, Sumitomo, Galaxy and Worldpay. More than 649 million dollars in USDC was bridged onto Arc in its opening week.

Around it, Circle has been assembling the rest. 

 reached 14.7 billion dollars in annualized volume with 175 enrolled financial institutions at the end of the second quarter. Circle Mint Europe issues both USDC and EURC from Paris under an ACPR licence. Circle's national trust bank charter received final approval from the United States Office of the Comptroller of the Currency on 10 July 2026. And on the agent side specifically, 99.3 percent of observed x402 transaction value settles in USDC.

That last number is the quiet one. When software pays software today, it overwhelmingly pays in Circle's dollar.






We have been on Arc since the first day of testnet, 28 October 2025, and we have never deployed to another chain.

Meridian is one connected loop, not a single product. Meridian Pay accepts payment in regulated digital dollars and euros. YieldVault puts idle balances into tokenized Treasury bills, with no fixed or guaranteed rate offered or implied. Payment Streaming pays by the second rather than at month end. And Firmata Protocol answers who is on the other side before the money moves.

Firmata composes ERC-8004 for agent identity and reputation with ERC-8183 for conditional escrow, settled over x402. It performs no anti-money-laundering or know-your-customer checks and replaces nobody's obligations. It publishes and verifies attestations about a counterparty, and it holds funds against a condition. That is the whole claim, and we will not inflate it.

Our stack runs on Arc testnet. Twenty contracts, each verified on Arcscan from the second of deploy. An independent security audit gates the move to mainnet, Soon, and we have not made it. We could have pushed the same bytecode to mainnet the week Arc launched and reported an address. We did not, and we would make the same call again.




Four things, because a piece built on somebody else's forecast should be honest about what a forecast is.



A prediction is not a rule. No G20 regulator has issued the guidance Lau describes. Three might. None might. The prediction could be early, late, or simply wrong, and the author would be the first to say so.

The report is a commissioned document. Mastercard chose the futurists and chose to publish the result. That does not make the analysis wrong. It means the questions were framed by a party with an interest in the answers, which is true of most industry research and is worth saying out loud.

We are not neutral. We build agent trust infrastructure. A dated forecast that agent trust infrastructure becomes a compliance requirement is a forecast that flatters our roadmap, and you should discount it accordingly. That is exactly why the provenance above is spelled out rather than summarized.

The report does not mention us. It does not mention Firmata, it does not mention Meridian, and nothing here should be read as suggesting otherwise. A global network named a category we work in. That is all, and it is enough.






In July, Know Your Agent became a category. We wrote about it then, and the argument was that most of what shipped into the category solved the easy half: identity without enforcement.

In September, the category got a countdown.

The gap between those two sentences is the entire opportunity. Categories reward whoever markets fastest. Deadlines reward whoever started earliest.

We started on 28 October 2025, the first day of Arc testnet, because the questions were already obvious. Who is this agent. Who authorized it. What happens when the job does not complete. Those questions did not become interesting because a card network published a report. The report is just the first time somebody attached a date to them.

🔗 

⏳ Know Your Agent Was a Category in July. In September, It Got a Deadline.

Home

Forum

Events

Content

Courses

Collections

People

Contributors

Albums

Groups

Chat

Channels

Not being used

Help

Join the global network of builders developing the Economic OS for the internet. We’re reimagining how value moves, creating the tools that make the economy work for everyone.

Arc House

What is your job title?

Join a global network of builders developing the Economic OS for the internet. We’re reimagining how value moves, creating the tools that make that economy work for everyone. 

Arc House | Build the economic OS for the internet with us

Copyright

Terms of Service

Privacy

Code of Conduct

Your Privacy Choices

It looks like there is no account associated with the email you provided. The event is for members only. Please complete the application to join our community first.

Thank you for your interest in joining us

Thanks for your interest in joining the event. However, your account is still pending. Once it's approved, you'll be able to register for the event. Please contact us via community@arc.io if you need support.

Pending Approval

Login to the The Arc Community Hub community to connect with others, attend community events, and more!

Login or Sign Up for even more

To unlock all parts of the community complete your profile using verifiable and real information.

Hey, complete your profile

It looks like there is no account associated with the email you provided. Please complete the application to join our community.

Thank you for your interest in joining Arc House.

Thanks for your interest in joining Arc House. We will review your request shortly and email you once it's approved. Meanwhile, please feel free to contact us via community@arc.io for any questions.

Accounts are unapproved when profile information cannot be verified or when activity patterns indicate spam, inorganic participation, or behavior that undermines trust in the program.

To strengthen your profile for future consideration, sign in with a verified LinkedIn account, use real and current profile information, and participate constructively in other community spaces.

Commercial customers should raise concerns through their support channel.

https://www.arc.network/community-guidelines