Fraudulent ChatGPT Pro charges spark alarm over unauthorized use of stolen card data

ChatGPT app icon is seen in this illustration taken on June 5. Reuters-Yonhap

Card companies and financial regulators in Korea are scrambling to respond to a wave of suspected fraud involving OpenAI's premium ChatGPT Pro subscription service, after hundreds of unauthorized charges were allegedly made using stolen card information.

Consumer complaints and testimonies have surged on Instagram and other social media platforms, as well as in local media reports, with users saying they were charged 299,000 won ($196) per month for ChatGPT Pro despite never subscribing to the service.

According industry data, a total of 1,368 ChatGPT Pro subscription payments worth about 400 million won were processed in Korea so far this month. Of those, 858 transactions, valued at around 250 million won, have been suspected of being fraudulent charges.

These fraudulent transactions were identified across nine domestic card issuers.

Cybersecurity experts said fraudsters likely used personal information obtained through phishing crimes or purchased on the dark web to create large numbers of ChatGPT accounts.

Unlike domestic online payments, some overseas transactions can be authorized using only a card number, expiration date and card verification code, making them more vulnerable to fraud.

OpenAI said the incident did not involve unauthorized charges generated by ChatGPT itself, but rather the misuse of stolen card information by third parties.

"We have confirmed that the payments were made using stolen card information without the holders' consent, and the affected payment methods were immediately disabled," the company said.

NICE Information and Telecommunication, a domestic electronic payment gateway, said it had canceled more than 700 transactions suspected of being fraudulent and is investigating the remaining cases.

The Financial Supervisory Service, the country's financial watchdog, has also reportedly asked card firms to strengthen monitoring through their fraud detection systems.