Your bank has likely used AI to watch your card payments for about 20 years. What is new is that attackers can use AI tools too, and the fakes they make are getting harder to tell from the real thing.
How AI guards payments today
Banks use machine learning to study transaction patterns and flag possible fraud. With 3D Secure 2, a card check system, the level of checking can adapt to a customer's behavior and history. A familiar purchase may go through with no extra steps. An unusual one, such as a purchase made from another country, may trigger verification.
These systems make mistakes. Real payments can still be flagged as suspicious. That is why human specialists need to keep monitoring and refining the models.
What is changing
The bigger shift is happening around the payment, not inside it. AI makes it easier to create convincing impersonations and deepfakes, which are fake voices, images or video. Firewalls and PCI DSS card security requirements remain essential. But businesses now face a growing volume of threats, and security teams increasingly need AI to detect and respond to AI-enabled attacks.
What is not known
The article gives no figures on how common these attacks are or how often real payments are wrongly flagged. It also does not name specific tools businesses should adopt.
The takeawayIf you run a business, keep your basic security in place, but expect AI-made impersonation attacks and the occasional wrongly flagged payment.