PayPal launches AI system to block scams before transactions complete

PayPal has launched a new AI-powered scam alert system that can intercept transactions before they’re completed, warning users about potential fraud in real-time. The system uses continually learning AI models to detect emerging scam patterns and provides dynamic warnings that vary based on risk levels, from simple alerts to complete payment blocks.

How it works: The AI system analyzes billions of data points to identify risk signals and adapts to new scam types without being specifically trained on them.

• PayPal’s models use “continually learning” technology that can detect similarities between known scams and new ones, allowing them to catch previously unseen fraud attempts.
• The system remembers new scam information for future detection efforts, creating a self-improving defense mechanism.
• “With continually learning models, our AI can understand similarities between a known and a new scam and help us detect the new one,” said Yigit Yildirim, SVP of global fraud prevention at PayPal.

Dynamic warning system: Unlike static fraud alerts, PayPal’s system provides different levels of intervention based on the confidence level of scam detection.

• High-confidence scam attempts trigger complete payment blocks to prevent money from being sent.
• Lower-risk transactions receive lighter interventions, such as safety verification prompts or recommended actions.
• The differentiated approach prevents “warning fatigue” that occurs when users repeatedly see the same generic alerts and begin ignoring them.

Rollout details: The scam alert system is being deployed across PayPal’s global network for Friends and Family transactions.

• The feature is available on the main PayPal platform worldwide and on Venmo in the United States.
• The system specifically targets person-to-person payments, which are common vectors for social engineering scams.

Why this matters: The launch comes as cybercriminals increasingly use generative AI to create more sophisticated scams, forcing payment platforms to evolve their defense strategies.

• “We need to be smarter and faster than the scammers, and we’re combating AI used by bad actors with more intelligent AI used for good,” Yildirim explained.
• The system addresses a key challenge in fraud prevention: balancing security with user experience by avoiding unnecessary friction for legitimate transactions.

The bigger picture: PayPal’s approach represents a shift toward proactive, AI-driven fraud prevention that can adapt to emerging threats rather than simply reacting to known patterns.

• Traditional fraud detection often relies on recognizing previously seen scam types, while PayPal’s system can identify new threats by analyzing behavioral patterns and transaction anomalies.
• The move positions PayPal to stay ahead in the “cat-and-mouse game” of cybersecurity as both attackers and defenders leverage increasingly sophisticated AI tools.