OpenAI's ChatGPT Pro subscriptions are linked to a wave of suspected fraudulent charges in South Korea involving unauthorized use of stolen card data. Card companies and financial regulators in South Korea are responding to the incidents.
What happened
Korea Times reports that a wave of suspected fraudulent charges linked to OpenAI's ChatGPT Pro subscription hit customers of nine South Korean card issuers. According to Korea Times, roughly 858 suspected fraudulent transactions with a combined value of approximately 250 million won were identified. The unauthorized charges correspond to the ChatGPT Pro tier, priced at 299,000 won (~$196/month), billed to card holders who had not enrolled in the service.
How it happened
Cybersecurity analysts cited by Korea Times assessed that the fraudsters likely used personal data acquired through phishing campaigns or purchased via dark web markets to create large numbers of ChatGPT accounts with stolen card credentials. Card companies and South Korean financial regulators are reported to be responding to identify and reverse the unauthorized charges.
Industry context
Editorial analysis: AI subscription services present an emerging target for card fraud, combining high monthly unit prices with automatic renewal mechanics that may delay victim detection. The $196/month price point for ChatGPT Pro is significantly higher than most streaming or software subscriptions, raising the per-transaction value for fraudsters. Payments teams and fraud-prevention vendors are expected to update detection models to flag AI subscription charge patterns originating from unfamiliar geographies or devices. The incident is regionally contained so far but illustrates a pattern likely to recur as premium AI tier pricing rises.
Scoring Rationale #
A regionally contained payment fraud incident involving AI subscription credentials. Relevant to payments and security practitioners as an emerging fraud pattern, but the story is a single-outlet regional report with limited broader AI/DS practitioner impact. Scores in the minor range.
Practice with real Payments data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card