Not for anything that matters yet. ChatGPT Wallet protects your card, but the store barely gets a say, and an agent's mistake counts as yours.
You can now tell ChatGPT to buy something for you, with your card. ChatGPT Wallet asks you to approve every purchase with Face ID or a fingerprint, and the store gets a stand-in card number for that one purchase, never yours. That was the part I expected to be risky, and it is the best-built part. The same day, OpenAI published a playbook for stores: its agent can already use their websites, they cannot add a confirmation step of their own, and an agent's mistake counts as the customer's. United Airlines has already said no, and its CEO called it another middleman. I run my own test for a middleman, and say what I would do as a shopper and as a store.
KEY TAKEAWAYS
Read the ChatGPT Wallet review screen as if you filled the cart yourself, because to the store you did.
Use an AI agent only for purchases you would be fine getting slightly wrong, like a reorder from a store you know.
If an agent's purchase fails, try the store's own website, because the store may have blocked the agent.
If you sell online, find the personal AI agents setting in your CDN or firewall and decide on purpose.
If you let AI agents in, tell your fraud provider about the extra signals OpenAI sends with them.
SHOW NOTES
ChatGPT Wallet Protects Your Card With Face ID And A Stand-In Card Number
Since 13 November 2026 you can tell ChatGPT to go and buy something for you, and it will, with your card. OpenAI's agents are called dots: ChatGPT doing a task for you in a browser in the cloud. A dot goes to the store's website, fills the cart and asks you before it pays. You see the items, the total and the card, and you approve with Face ID or a fingerprint, every time.
Then Visa, Mastercard or American Express makes a stand-in card, with its own number, security code and expiry date, good for that one purchase. The dot uses it at the store's normal checkout and the store's own payment company charges it. The store never sees your card number. For now it is ChatGPT Pro only, in the US, through dots, and OpenAI has not named a store that has been paid through it.
OpenAI's Playbook Gives Stores Little Control Over Its Shopping Agent
OpenAI's last try at shopping, Instant Checkout, asked stores to build something to sell inside ChatGPT, and few did. This time the agent goes to the store. On the same day as the wallet, OpenAI published a playbook for stores: upload your products, let OpenAI's agent through your security, and talk to whoever checks your orders for fraud. It also says the agent can already use a store's website, with no permission needed.
Its answers to stores' questions say how little control they have. There is no feature yet for a store to add a confirmation step of its own. The only check is the shopper's own approval. An exception for the agent covers the whole website, and the store's existing page-level controls stay in place. Customers are ultimately responsible for what the agent does. And stores that block access may lose customers who turn elsewhere.
By My Test, OpenAI Is A Middleman, And It Calls That Buying Directly
United Airlines lets AI agents show its flights but not book them, and its CEO, Scott Kirby, said he does not want another middleman. OpenAI's wallet page says shoppers transact directly with businesses, which keep their customer relationships.
My test for a middleman: does it take a cut, collect data, or anything else? OpenAI told CNBC it is not taking a cut to start, and it does not process the payment. But the conversation where you choose what to buy happens in ChatGPT, stores that upload products send OpenAI their prices and stock, and that product list can also feed ChatGPT's ads, if the store switches that on. Apple Pay sits in the middle too, but only carries the payment. This one makes the choice. That is why, for anything that matters, my answer is not yet.
ChatGPT Wallet is OpenAI's payment wallet for its AI agents, called dots. You add a card once, and when a dot buys something for you it asks you to approve the purchase with Face ID or a fingerprint. It launched on 13 November 2026 for ChatGPT Pro subscribers in the US.
Is ChatGPT Wallet safe?
The payment part is well designed. Every purchase needs your approval with Face ID or a fingerprint, and the store receives a stand-in card number that works for that one purchase, never your real card number. The bigger risk is that if the agent buys the wrong thing, the store treats it as your own mistake.
Can stores block ChatGPT agents?
Yes. Stores can block OpenAI's agent in their CDN or firewall settings. On Cloudflare it appears as dots, OpenAI, under Personal Agents, and it was set to Allow by default on No Hacks' website. United Airlines lets agents show its flights but not book them. OpenAI says stores that block agents may lose customers.
Does OpenAI take a cut of ChatGPT Wallet purchases?
Not for now. OpenAI told CNBC it is not taking a cut of transactions to start, and the payment is processed by the store's own payment company, not by OpenAI. OpenAI's instructions for stores do not mention a fee.
WEBVTT
00:00:00.000 --> 00:00:28.000
You can now tell ChatGPT to go and buy something for you, and it will with your card. The feature's called ChatGPT Wallet. OpenAI announced it on November the 13th, Friday the 13th. On that same day, they published instructions for the stores, and one of the first things they say to the stores is that ChatGPT can already use their websites, and they don't need to do anything to put it on a list of approved visitors.
00:00:28.000 --> 00:00:43.000
So the store does not get asked. ChatGPT will show up either way. This is No Hacks. AI is changing the way you use the internet, and nobody has asked you about it. So I check what AI is actually doing to it, the good, the bad, and the pointless.
00:00:43.000 --> 00:01:11.000
Should you let ChatGPT buy things for you yet? I expected the wallet, the money part, to be the risky part, and I was wrong about that. The wallet is good. The problem is what's being built around the wallet. Step by step, OpenAI is putting itself between you and the stores you buy from. That's a middleman, or as OpenAI calls it, buying directly. AI shopping is a new way to discover and choose what to buy without ever looking at a store's website.
00:01:11.000 --> 00:01:33.000
You ask, the AI finds, compares, and now buys for you. And if you never look at the store yourself, whoever does the looking for you matters a lot. Right now, almost nobody shops this way. DoorDash, one of the biggest food delivery apps in the US, says outside AI agents have placed maybe several hundred orders in total.
00:01:33.000 --> 00:01:40.000
In total! And still, every few weeks, somebody builds another piece of this puzzle. It's a bit like self-checkout when it first showed up.
00:01:40.000 --> 00:01:45.000
Supermarkets put the machines in, most people still went to the cashier.
00:01:45.000 --> 00:01:47.000
And the machines kept getting better anyway.
00:01:47.000 --> 00:01:51.000
And the rules for it are being written right now.
00:01:51.000 --> 00:02:10.000
OpenAI's agents are called dots. A dot is ChatGPT going off and doing a task for you in a browser, in the cloud, not on your phone or your computer. It reads pages, it clicks buttons, fills in forms. A purchase goes like this. You add your card once. The dot goes to the store's website and fills the cart like you would.
00:02:10.000 --> 00:02:31.000
Before it pays, it asks you. You see the items, the total, and a card, and you approve it with your face ID or your fingerprint every single time. And honestly, that part is really well-designed. And then Visa or Mastercard or American Express makes a stand-in card, a different card number with its own security code, its own expiry date.
00:02:31.000 --> 00:02:42.000
It's the same idea as Apple Pay, except this one is good for that one purchase only. The dot uses that stand-in at the store's normal checkout, and the store's payment company charges it.
00:02:42.000 --> 00:03:02.000
The money comes off your real card, and the store never sees your card number. For now, this is only ChatGPT Pro, OpenAI's expensive tier, in the US, and only through dots. It started on November thirteenth. So there's no number yet on how many people have used it, and OpenAI has not named a single store that's been paid through it.
00:03:02.000 --> 00:03:24.000
And the pro part, I wonder if that's a look at the future. Google's newest model in its AI search is only on its most expensive plan. Meta says its AI only makes phone calls for you if you pay. The AI that actually does useful things shows up first for the people who pay extra. Maybe that will change in the future, maybe it will not.
00:03:24.000 --> 00:03:41.000
Either way, your card is fine. But the buying itself, it happens on the store's website, and that's new because this is not OpenAI's first try at shopping. The last one was called Instant Checkout. You bought inside ChatGPT, and the store had to build something to be there.
00:03:41.000 --> 00:04:03.000
Six months after it was announced, about 30 Shopify stores had, according to Forrester, a research firm, used it. And then in March this year, OpenAI dropped it. Its page now says it is moving away from a standalone instant checkout experience, and that was the right call. Last time, OpenAI asked the stores to come to it, and very few did.
00:04:03.000 --> 00:04:06.000
This time, OpenAI is going to the stores.
00:04:06.000 --> 00:04:27.000
and those instructions OpenAI has for stores, this is what they're for. OpenAI calls them a playbook. You upload your products with prices and stock. You make sure your security lets OpenAI's agent through. You talk to the company that checks your orders for fraud. Look, uploading products is not new, but now it's one of the first things OpenAI asks for.
00:04:27.000 --> 00:04:52.000
Back in June, I had Malte Landwehr from Peec AI on the show. Peec is a company that tracks what AI tools recommend. His team, together with an independent researcher looked at ChatGPT's shopping results. The products, the prices, the ratings, they matched Google Shopping. ChatGPT was showing you Google's shopping data. Now OpenAI is asking every store for its products directly.
00:04:52.000 --> 00:05:17.000
In the playbook, there's a list of questions stores ask with OpenAI's answers. So a store might ask, can I require that a human confirm the checkout? OpenAI says it does not have features for that yet, and it adds that across all its users, it requires human review before any purchase like that. So the human who confirms this is you if you're buying. The store cannot ask for anything more.
00:05:17.000 --> 00:05:34.000
A store might ask, can I keep agents away from certain pages or products? The answer, the exception that lets the agent in can apply site-wide, and whatever page level controls the store already has stay in place. So there's no special switch to keep the agent away for some products.
00:05:34.000 --> 00:05:38.000
A store might ask, who handles payment or order errors?
00:05:38.000 --> 00:06:00.000
OpenAI answers, "The customers are ultimately responsible for the actions they take." So if the agent orders the wrong size to the store, that's you ordering the wrong size. And then, then there's fraud. Fraud tools are built to stop bots, and OpenAI's agent is a bot, so purchases can fail because of this.
00:06:00.000 --> 00:06:10.000
OpenAI's fix is to start sending the store extra information about you, your IP address as OpenAI sees it, your country, your time zone, your language.
00:06:10.000 --> 00:06:16.000
The store will see most of that if you visited yourself, but now the agent carries all that data for you.
00:06:16.000 --> 00:06:32.000
And if the store doesn't like any of this, well, there's a question for that, too. How will this affect my relationship with my customers? OpenAI's answer is that the websites that block access may lose users or customers who turn to other sources to make decisions.
00:06:32.000 --> 00:06:33.000
This is important.
00:06:33.000 --> 00:06:39.000
OpenAI also says that a billion people use ChatGPT every week. That's their own number.
00:06:39.000 --> 00:06:49.000
Some businesses are saying no to this kind of setup anyway. This week, United Airlines said AI agents can show you United flights, but they cannot book them for you.
00:06:49.000 --> 00:07:09.000
United's reasons are privacy, purchases people did not mean to make, and federal rules. When you book a flight, you have to confirm things. Like that question, do you have anything dangerous in your bags? An agent cannot do that for you. And then United CEO Scott Kirby gave the other reason at an event on Tuesday.
00:07:09.000 --> 00:07:29.000
According to The Wall Street Journal, he said, "To be honest, I don't want to create another middleman." The way I read that, for an airline, the middleman is whoever you book through. They get your email, the upgrade offer, the next trip. An AI agent that books for you would be a brand-new one sitting on your phone.
00:07:29.000 --> 00:07:40.000
On the wallet page, OpenAI says shoppers transact directly with businesses which keep their customer relationships and control over pricing, fulfillment, returns, and customer service.
00:07:40.000 --> 00:07:53.000
And I don't know who's right here, OpenAI or United, but obviously this is a conflict because OpenAI never says they're not the middleman. I think OpenAI is trying to get in as a middleman without addressing the fact that they're a middleman.
00:07:53.000 --> 00:08:03.000
And my test for what a middleman is, is simple. Do they take a cut? Do they collect the data or anything else? If yes, they're a middleman.
00:08:03.000 --> 00:08:05.000
So OpenAI, a cut?
00:08:05.000 --> 00:08:14.000
OpenAI told CNBC it's not taking a cut to begin with, and it doesn't even process the payments. The store's payment company does.
00:08:14.000 --> 00:08:30.000
Data? That's a different story. A store that uploads its products sends OpenAI its prices and stock, And the conversation where you decided what to buy happens in ChatGPT, not on the store's website. They can't even see your page views in analytics.
00:08:30.000 --> 00:08:38.000
Anything else? Also, yes, the product list the store uploads can also feed ChatGPT's ads if the store switches that on.
00:08:38.000 --> 00:08:46.000
That's a middleman. On that same page, OpenAI calls this buying directly from the business.
00:08:46.000 --> 00:09:02.000
That's directly through OpenAI's agent paying with OpenAI's wallet in OpenAI's browser. I'm not sure OpenAI is trying to avoid the word, but calling that directly is not honest.
00:09:02.000 --> 00:09:22.000
And a middleman is not necessarily bad. Apple Pay sits in the middle too, but Apple Pay only carries the payment. You still pick the store. You look at the product. You decide what to buy. Here, the AI does the looking and the choosing inside ChatGPT, And the store sees the agent and the order, never the conversation that chose it.
00:09:22.000 --> 00:09:32.000
That's the difference. Apple Pay is a tool you use. This is a tool that does things for you, and making the choice is one of the things it does.
00:09:32.000 --> 00:09:35.000
So what does having another middleman cost you?
00:09:35.000 --> 00:09:56.000
If the agent gets it wrong, the mistake is yours with a store that never agreed to the agent in the first place. And the choice of what to buy and where happens somewhere the store cannot see and you cannot check. And that's why for me, this is a not yet, not for anything that matters.
00:09:56.000 --> 00:10:00.000
Most people cannot use it yet anyway, but when it gets to you, I would do three things.
00:10:00.000 --> 00:10:16.000
The one moment you do control is the review screen. The items, the total, the cart, that's the last thing between you and the order. Read it the way you would read it if you'd filled the cart yourself, because for the store, that's exactly what happened.
00:10:16.000 --> 00:10:28.000
Use it for things you'd be fine getting slightly wrong. A reorder from a store you already know is fine. A flight isn't, and neither is anything you'd need to return or argue about.
00:10:28.000 --> 00:10:37.000
And if it fails, it might not be broken. The store might have said no. The store's website still works.
00:10:37.000 --> 00:10:42.000
If you happen to sell anything online, there's a setting that has probably decided this for you.
00:10:42.000 --> 00:11:00.000
I checked my website security dashboard on Cloudflare, which sits in front of a lot of websites. There's a section called Personal Agents. It lists four AI agents by name, Claude Duet from Anthropic, dots from OpenAI, Gemini Spark from Google, Muse from Meta, and there's a fifth one coming soon. Each one is a dropdown.
00:11:00.000 --> 00:11:22.000
All four were set to allow for me by default. I never touched those settings. That's how they came. And for agents, Cloudflare even marks Allow as recommended. On my website, on No Hacks website, that is completely fine. I don't sell anything there. But if you have a store, that dropdown is the kind of decision United Airlines made.
00:11:22.000 --> 00:11:54.000
The difference is that they made it on purpose. If your website is not on Cloudflare, the same setting usually sits with whoever runs your security layer. If I had a store, it would depend on the store. If the orders are small and returns are easy, I'd let them in, and then I'd tell my fraud provider about the extra information OpenAI sends so the agent's orders do not get declined. If a wrong order is expensive or there's something the customer has to confirm themselves, I'd do the same thing United Airlines did, at least for now.
00:11:54.000 --> 00:12:09.000
The product upload is a separate decision. It helps ChatGPT show your products. It also puts your catalog with OpenAI. If ChatGPT sends you customers, I would do that. If it doesn't, it's probably okay to wait.
00:12:09.000 --> 00:12:25.000
On November 10th, the CEO of an airline looked at AI that books for you and called it a middleman. Three days later, OpenAI launched AI that buys for you and called that buying directly. One of those is wrong.
00:12:25.000 --> 00:12:45.000
Thank you for tuning in. I would like to remind you that every week I have a newsletter at nohacks.co/subscribe, And as far as this episode is concerned, please consider rating, reviewing, and sharing it with someone who would benefit from it. I will talk to you next week.
Mark Zuckerberg says Meta's AI agent, Muse, needs to be discreet. The same month, some of the calls it made to businesses were placed by people in a call center. I went through what he told Joanna Stern and Alex Heath about Muse's privacy and security and sorted every claim by tense. What exists today, only Meta has checked. The protection from Met...
On August 5 Shopify switched on a second way into every store on its platform, built for AI agents instead of people. Nothing to install, no merchant asked. I opened three unrelated stores, Allbirds, Brooklinen and Partake Foods, and asked each what it could do for a machine. All three answered with the same roughly 800 words, identical character f...
I sat down with Malte Landwehr, who left VP of SEO at Idealo to become CPO and CMO at Peec AI, the platform that tracks what ChatGPT, Claude, Gemini, and Google AI Overviews actually cite. We open on the strangest finding of the year. GummySearch, a Reddit analytics tool that shut down last November, now sits behind about 0.1% of all ChatGPT citati...
Practical strategies for making your website work for AI agents and the humans using AI to find you. Once a week you get the new articles, the latest podcast episode, and a few links worth keeping.
