Zoli Rutter, a Metro Bank customer from Sussex, is fighting for a refund after falling victim to an AI-linked fraud that drained £14,244 from his account. The scam involved criminals purchasing credits to use Claude, the AI chatbot developed by Anthropic, using Rutter's debit card details.
According to the Guardian, the fraudsters exploited a situation where Rutter had previously made legitimate payments to Anthropic, providing his card details. When Metro Bank alerted Rutter about an attempted transaction, he replied "no," but the criminals still managed to process a series of subsequent payments ranging from £90 to £200 each. Rutter had been paying approximately £15 a month to use the chatbot for several months before the incident.
How the scam unfolded
The fraudsters used the Metro Bank debit card linked to Rutter's Anthropic account to make unauthorized purchases of credits on the platform. Despite the bank's alert and Rutter's negative response, the criminals were able to continue making transactions. A Metro Bank spokesperson explained that due to the "complex nature of the fraud, regretfully some payments were processed" before the bank fully blocked the card.
The bank told the Guardian: "This is a concerning case involving a third party that the customer had made previous genuine payments to and provided card details [to], which were then used to make a series of purchases without Mr Rutter's knowledge."
Metro Bank's response and refund
Metro Bank has acknowledged the distress caused to Rutter, stating: "We recognise how worrying and frustrating [the] experience has been for Mr Rutter. Metro Bank has refunded Mr Rutter while we progress a chargeback dispute request with Anthropic to refund the money that had been taken."
The bank's decision to refund Rutter comes as a relief, but it highlights the growing threat of AI-related fraud. In this case, the criminals exploited the trust established between the customer and a legitimate service provider, a tactic that is becoming increasingly common.
Broader context: AI model extraction concerns
This incident occurs against a backdrop of heightened concerns about AI security. In June, Anthropic accused Chinese e-commerce and technology firm Alibaba of "brazenly" and "illicitly" extracting its Claude AI model's capabilities. In a letter to two members of the US Congress, the San Francisco-based company said operators linked to Alibaba carried out almost 29 million exchanges with Claude using thousands of fraudulent accounts, calling it the largest extraction campaign of its kind.
Anthropic urged Congress to penalise companies behind such attacks and to ramp up measures to prevent US technology from being stolen. This case underscores the vulnerabilities in AI platforms and the need for stronger safeguards to protect users and companies alike.
What consumers should do
For UK consumers, this incident serves as a reminder to monitor bank statements regularly and respond promptly to any suspicious activity. Banks typically offer protection against unauthorised transactions, but it is crucial to report any anomalies immediately. In cases where fraud involves third-party services, customers should be aware that refunds may depend on the bank's discretion and the outcome of chargeback disputes.
Metro Bank's decision to refund Rutter while pursuing a chargeback with Anthropic demonstrates a customer-centric approach, but it also highlights the complexities of resolving such fraud cases. As AI services become more integrated into daily life, both financial institutions and tech companies must collaborate to enhance security measures and prevent similar incidents.
