Cybersecurity
September 23, 2026
Fraud is entering a new era. AI is giving fraudsters new ways to find vulnerabilities, adapt their tactics and scale attacks. At the same time, money is moving across more channels, including increasingly autonomous experiences. This gives consumers more convenience, but it also gives fraudsters more places to hide and defenders less time to respond.
The scale of the challenge is significant: Last year, scam losses globally were estimated at $442 billion, and generative AI could fuel $40 billion in U.S. fraud losses by 2027. The need to identify risk before it becomes a fraudulent transaction or scam payment has never been more urgent.
For years, fraud prevention has focused largely on assessing individual transactions and responding to known threats. While AI has been driving up cybersecurity risks, this same technology is increasingly being used to fend off these attacks. AI creates an opportunity to connect intelligence across the digital ecosystem, understand how fraud develops and intervene sooner — shifting from reacting to fraud to anticipating it.
Fraud rarely reveals itself through a single transaction. It emerges through patterns that can be difficult to recognize when each signal is viewed in isolation.
Imagine a consumer who usually shops in Seattle and Portland. Within minutes, several high-value online purchases appear from unfamiliar merchants in another country. AI can examine the broader pattern — including location, purchase frequency, value, spending history and device information — to identify the activity as high risk before more fraud occurs. At Mastercard, we use generative AI to analyze complex transaction patterns and identify subtle signals that traditional rules may miss, helping issuers detect more fraud while reducing false declines.
The same principle applies to complex fraud networks. AI can help map connections among accounts, devices, transactions and other digital activity to reveal coordinated fraud that may be hidden when each signal is viewed alone.
Fraud may surface at the point of transaction, but it often begins much earlier: a compromised password offered for sale online, a fake storefront created to deceive shoppers or a phishing campaign collecting information criminals later use to access accounts or make fraudulent purchases.
By bringing together cyber threat intelligence and payments intelligence, AI can connect these early warning signs and help identify where risk is taking shape. The same approach can help identify merchant risk earlier. By analyzing digital footprints, onboarding signals and transaction activity, AI can detect potentially fraudulent or compromised merchants before they create broader ecosystem exposure.
Scams are one of the clearest examples of why fraud prevention must evolve. Scam payments are often authorized by the victim, who believes they are paying a legitimate person or business. That makes them harder to detect using traditional fraud controls.
By connecting behavioral, account, and transaction signals, we are using AI to identify manipulation earlier. Instead of asking only, “Is this transaction fraudulent?” AI can help us answer: “Is this consumer being manipulated into sending money?,” giving financial institutions an opportunity to intervene before harm occurs.
The shift towards prevention will become even more important as AI agents begin acting on behalf of people and businesses. Trusted and malicious activity may both move quickly, operate autonomously, and look unfamiliar compared with the patterns we recognize today.
In an agentic economy, a consumer may ask an AI agent to book travel, compare products or complete a purchase. That creates new questions: How does the ecosystem know that the transaction reflects a consumer’s intent, that the agent is authorized to act, and that the payment is secure?
As agentic commerce grows, fraud prevention will need to verify not only identity, but also intent and authorization. Success will depend on connecting intelligence across the ecosystem to identify risk earlier and build trust in new ways to transact.
Most people will never see our team’s intelligence working behind the scenes. But they will experience the benefits every day through safer transactions, stronger protection from scams and greater confidence in new forms of digital commerce. That is the promise of AI: not simply detecting fraud faster but stopping it before it starts.
