A businessman in Sussex watched £14,244 drain out of his Metro Bank account. The money went on credits for Claude, the AI chatbot he already paid to use. Two things failed at once. Criminals had turned a chatbot into a way to cash out stolen cards. And his bank spotted the first fraudulent charge, warned him, then let more than £14,000 more go through.

Zoli Rutter had been paying about £15 a month for Claude for a few months, he told Guardian Money. He used it to track and analyse business invoices. The fraudsters used the Metro debit card linked to his Anthropic account to buy credits on the platform.

Caught, then waved through

On 19 June, Metro spotted a suspicious transaction for £90.90. It texted Rutter to ask whether he had authorised it. He texted back to say he had not. He then got another message: Metro would freeze his account. In the event, the bank blocked only that one transaction, not the account.

The fraudulent charges kept coming, each between £90 and £200. Metro blocked some and missed many. It froze the card the day after the scam began, but by then more than £14,000 had gone. The bank blamed the “complex nature of the fraud.” It had processed some payments, it said, before it could block the card fully.

A chatbot as a cash register

The mechanics are what make this new. A coordinated gang used Rutter’s card details to buy Claude credits, according to Anthropic. The company said it had no evidence the details came from its own systems. It had also banned the user behind the fraud. Anyone wrongly charged should contact its support, it added, and it will refund the charges.

It was not a one-off. In May, a US Claude user told Guardian Money that fraudsters had bought gift cards for the chatbot with his financial details. Others posted similar accounts on Reddit. AI credits are becoming a target in their own right. It is part of a wider AI-era fraud economy that banks are racing to keep up with.

Who pays

The refund became a tangle. After Guardian Money got in touch, Metro temporarily refunded Rutter while it chased a chargeback from Anthropic. Anthropic has since refunded him directly, so the bank will now claw back its temporary payment. Metro told Rutter in a letter that he had not received the service it expects. It offered him £300 in compensation.

Rutter says he intends to complain to the Financial Ombudsman Service. Metro says it has taken steps to prevent delays in blocking cards in future.

The case lands in a busy month for Claude. Last week the FCA appeared to endorse the chatbot, agreeing to give firms in its high-tech sandbox access to it. This week it emerged that some people’s Claude conversations were publicly visible online. Anthropic has separately moved to verify who its users are.

The underlying problem is simpler than any of it. As more people save a card to pay for AI, that card becomes a target. The safeguards on both sides are still catching up.

Get the TNW newsletter

Get the most important tech news in your inbox each week.