The concept of *agentic commerce* is emerging as a significant development in digital payments. Instead of relying solely on instant payments or stablecoins, this approach uses AI-powered software to handle transactions on behalf of users. Mastercard and Visa are leading the charge, enabling these trusted agents to initiate payments within user-defined parameters. This shift has the potential to streamline transactions significantly, allowing AI to manage everyday purchases based on user preferences, available balances, and promotional offers. However, the move to agentic commerce introduces new risks. The focus of fraud prevention will shift from individuals to the software itself. Criminals may target AI agents with tactics like injecting malicious instructions or stealing cryptographic keys. Unlike traditional fraud, which often exploits human error, these threats will require a different detection strategy. It’s essential to monitor the behavior of software agents closely, looking for anomalies in transaction patterns rather than relying on personal verification methods. Security measures are being built into the agentic commerce framework. Each agent will possess a verified identity and a cryptographic key, ensuring that transactions are authorized and limited in scope. If an agent acts outside its defined limits, the transaction will be denied. This system aims to enhance transparency, providing clear data on who initiated payments and under what conditions. With the right data points, fraud prevention can adapt to this new landscape, leveraging agent-specific information to enhance risk assessment. As agentic commerce evolves, it represents a significant shift in how transactions are conducted. With billions of potential transactions on the table, the implications for payment systems and fraud prevention are substantial. Understanding these changes is crucial for banks, merchants, and the broader financial ecosystem, as they will need to adapt to new technologies and the associated risks.