Advance of Agentic Artificial Intelligence

Last week saw a significant milestone in agentic artificial intelligence as Mastercard showcased an AI agent during the India AI Impact Summit 2026. In this demonstration, the AI was able to locate and purchase a product without requiring user intervention or manual entry of payment details.

Security in Agentic Commerce

The primary focus of Mastercard’s demonstration centered on security. The transaction leveraged the Agent Pay framework to ensure both user and AI agent authentication before authorizing payments, highlighting the system’s ability to meet financial transaction standards.

Security Considerations

These stringent measures are particularly crucial as agentic systems broaden their scope into B2B applications. In these scenarios, transaction volumes are significantly higher and governance requirements more complex.

Mastercard’s Approach

Nitendra Rajput, Senior Vice President and Head of Mastercard AI Garage, emphasized the importance of secure automation and robust fraud prevention. He underscored that the payment architecture includes strict safeguards and user control mechanisms.

Future Opportunities

The broader commercial implications are substantial. According to Deloitte, agentic AI could achieve $17.5 trillion in gross merchandise value by 2030, with about $15 trillion coming from B2B transactions and the remainder from B2C purchases.

Role of Technology in Agentic Commerce

According to Don Apgar, Director of Merchant Payments at Javelin Strategy & Research, Visa and Mastercard’s efforts are pivotal steps toward building the infrastructure necessary for agentic commerce. He noted that while shopping bots may enhance personal convenience, the initial practical applications are likely in enterprise B2B purchasing tasks.