Metro Bank customer Zoli Rutter is fighting for a £14,244 refund after AI-linked fraud involving the Claude chatbot. The scam, which exploited his debit card linked to an Anthropic account, highlights the growing risk of AI-powered financial crime. In this article, we break down what happened, how to protect yourself, and your rights when banks fail to stop fraud.
How the AI Fraud Happened
Zoli Rutter, a businessman from Sussex, had been using Claude, an AI chatbot, for a few months to track and analyze business invoices. He paid about £15 a month for the service. Fraudsters obtained his Metro Bank debit card details and used them to buy credits on the Anthropic platform, making multiple transactions between £90 and £200.
Get the #1 Wireless Door Camera
REOLINK Bestseller: 2K Weatherproof Video Doorbell, No Monthly Fees.
On 19 June, Metro Bank spotted a suspicious transaction of £90.90 and texted Rutter to ask if he authorized it. He replied "no," and the bank said his account would be frozen. However, only that single transaction was blocked, not his entire account. Subsequent fraudulent transactions continued, with many going through despite some being blocked. The card was only frozen the following day.
Alarming Pattern of AI-Linked Fraud
This case is not isolated. In May, a US-based Claude user reported that gift cards for the chatbot were bought using his financial details. Similar stories have surfaced on Reddit, indicating a broader pattern of fraud involving AI platforms. The Financial Conduct Authority (FCA) recently endorsed Claude for its AI sandbox initiative, but this incident shows the risks associated with AI services.
What You Can Do to Protect Yourself
- Use strong, unique passwords for AI accounts and enable two-factor authentication.
- Monitor bank statements regularly for unauthorized transactions.
- Set up transaction alerts and consider using virtual cards for online subscriptions.
- If you receive a suspicious text from your bank, call the official number to verify, not the number in the text.
Your Rights When Banks Fail to Stop Fraud
Under UK regulations, banks must reimburse unauthorized transactions unless they can prove you were grossly negligent. In Rutter's case, he promptly replied "no" to the bank's text, yet the bank failed to freeze the account immediately. This could be a clear case of bank liability.
According to the Payment Services Regulations 2017, you are entitled to a refund for unauthorized transactions if you report them promptly. However, banks may argue that you shared your card details or failed to protect them. In this situation, Rutter did not share details; the fraudsters obtained them through other means.
Comparison: Bank Response in Fraud Cases
| Action Taken | Metro Bank (This Case) | Recommended Best Practice |
|---|---|---|
| Initial alert | Text about one transaction | Freeze entire account immediately |
| Customer response | Replied "no" | Confirm and freeze all activity |
| Subsequent transactions | Some blocked, many passed | All blocked until verification |
| Card freeze | Next day | Immediate |
Steps to Take If You Are a Victim
- Contact your bank immediately and request a full account freeze.
- Report the fraud to Action Fraud (UK) or your local authority.
- Keep all communication records and transaction details.
- File a formal complaint with your bank and escalate to the Financial Ombudsman if unresolved.
Rutter's fight for a refund is ongoing, but this case serves as a warning. Banks must act swiftly when customers flag suspicious activity. As AI tools become more integrated into daily life, both users and financial institutions need to adopt stronger security measures.