\u201cScalability will not be challenged by technology. It\u2019s going to be challenged by trust.\u201d<\/p>\n
In a recent conversation with BetaKit, Reiff revealed how Mastercard is adapting that layer of commerce for a new kind of interaction, where AI can take an interaction all the way from decision to purchase.\u00a0<\/p>\n
\u201cIt\u2019s really the next evolution of payments,\u201d Reiff said.\u00a0<\/p>\n
Over two decades ago, the world first experienced what it meant to shop online. \u201cThis is that next evolution,\u201d he added, \u201cmoving from online into the world of AI.\u201d<\/p>\n
As it stands today, Reiff described generative AI as largely confined to the research phase of a shopper\u2019s experience. It can give people a place to start and help narrow down options, but it stops short of execution.<\/p>\n
\u201cGenerative AI gives you information, but I still have to go to a store, I have to go buy something online,\u201d Reiff said. \u201cAgentic commerce closes that loop.\u201d<\/p>\n
Instead of moving from recommendation to check-out across multiple tabs and steps, the same system that helps a user decide what to buy can carry that intent through to purchase, completing the transaction.\u00a0<\/p>\n
Reiff sees the first use cases will come in purchases that are common but time-consuming, like planning travel or reordering household supplies. These aren\u2019t decisions most people struggle with, but they take time because the work is in coordinating the details, checking all the available options, and following through. Reiff believes agentic commerce can absorb much of that work.<\/p>\n
\u201cIt\u2019s really about putting time back in the hands of users and consumers,\u201d he said.<\/p>\n
When planning a vacation abroad, for example, a user could tell an AI agent they need to book flights, hotels, excursions and a car rental. The agent can then give them options based on personalized factors like their travel dates, airline loyalty, and hotel preferences. From there, the user can authorize those purchases directly through the AI agent.<\/p>\n
With agentic commerce, humans are still providing consent for the purchase, but with the help of AI, they no longer have to spend hours piecing the whole process together.<\/p>\n
\u201cHumans are setting the control and the authorization of everything,\u201d Reiff added. \u201cIt\u2019s, as we say: AI-assisted, human-led.\u201d<\/p>\n
At the inaugural Mastercard Innovation Forum in Canada in April<\/a>, over 500 industry leaders and innovators gathered to explore how Canada can advance what is possible in the digital economy, cybersecurity and payments technology \u2013 including agentic commerce. The forum\u2019s theme, Forward, Together, reflected a clear consensus that the next era will be led by organizations that lean into industry collaboration and turn data into an advantage responsibly, reliably, and in real time.<\/p>\n As agentic systems become more prominent for users and businesses around the world, Mastercard\u2019s Chief Innovation Officer, Ken Moore, led a key conversation about agentic commerce at the Forum.<\/p>\n \u201c[When] you give permission to an agent to act on your behalf, you\u2019re actually creating a new threat surface, a new threat vector,\u201d Moore said at the forum. \u201cBut we should also understand that in that threat, there are also opportunities.\u201d<\/p>\n He noted that AI can parse signals, spot patterns that humans can\u2019t see, respond in real time, and perhaps most importantly, take action and stop threats. But AI\u2019s capabilities alone aren\u2019t what will determine adoption.\u00a0<\/p>\n \u201cThe technologies exist today,\u201d Moore added. \u201cReally, the challenge is more about engineering trust into the ecosystem.\u201d<\/p>\n Moore called trust \u201cthe currency of innovation,\u201d and according to Reiff, winning the hearts and minds of users and businesses depends on whether they feel confident that the systems will do exactly what they expect.<\/p>\n \u201cTrust is the fundamental component,\u201d Reiff continued. \u201cScalability will not be challenged by technology. It\u2019s going to be challenged by trust.\u201d<\/p>\n \u201cThe technologies exist today. Really, the challenge is more about engineering trust into the ecosystem.\u201d<\/p>\n
![\"\"](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/05\/mastercard-scaled.png\" "\\"mastercard")
<\/p>\n