Visa sounds the alarm on AI fraud

Visa’s latest threat report highlights how AI is transforming cybercrime in banking, making fraud easier and faster for criminals while pushing financial institutions to adopt new security measures. Michael Jabbara, Visa’s senior vice president for Payment Ecosystem Risk and Control, stated that AI has lowered the technical skills required for sophisticated fraud, allowing attackers to automate and refine tactics in near real-time. Traditional fraud detection methods, relying on static rules, are increasingly ineffective against AI-driven attacks, which can now execute transactions and mimic legitimate user behavior autonomously. Generative AI and agentic systems are the biggest threats, according to industry experts. Jim Mortensen, a strategic advisor at Datos Insights, noted that 89% of financial institutions report deepfakes and AI as major drivers of payment scams. These tools enable criminals to create personalized phishing campaigns, voice impersonation scams, and synthetic identities at scale. Agentic AI can autonomously adapt to failed attempts, making it harder for fraud teams to distinguish between legitimate and malicious activity. The shift in fraud detection now focuses on identifying AI capabilities rather than just user identity. Real-time payments further complicate fraud prevention, as criminals exploit compressed timelines to move funds before legacy systems can respond. Datos Insights found that 58% of financial crime leaders cite inadequate detection capabilities for instant payments as a critical issue. Money-laundering schemes that once took weeks can now unfold in hours, as AI automates layering techniques across institutions. Top U.S. banks, including Bank of America, Citi, Goldman Sachs, Morgan Stanley, and Wells Fargo, recently met with Anthropic to discuss AI-driven fraud risks, particularly deepfake threats. Visa’s report emphasizes that the most dangerous fraud involves transactions appearing entirely legitimate, as customers unknowingly authorize fraudulent payments. The rapid rise of AI-enabled social engineering is forcing banks to invest in unified threat intelligence platforms to stay ahead of evolving attack methods.