Microsoft’s strategy centers on moving intelligence directly into frontline environments, where AI systems operate continuously rather than responding to discrete prompts. The company frames this transition as “frontier transformation,” describing retailers that adopt agentic systems as “frontier firms” that embed adaptive intelligence across physical stores, cloud platforms and enterprise software.
Microsoft’s thesis is that retailers face a widening gap between customer expectations and operational capacity. Agentic AI, the company argues, can close that gap by allowing systems to monitor conditions, make decisions, and execute tasks in real time while escalating exceptions to human workers.
Agentic Robots Take on Operations
Microsoft’s most concrete example is ADAM, a robotic system developed with Richtech Robotics and powered by Azure AI. Originally designed as a beverage-serving robot, ADAM now operates as a conversational, context-aware assistant. Microsoft says the robot adjusts drink recommendations based on factors such as time of day, weather and promotions, while responding to customer requests like sweetness or ice preferences in natural language.
ADAM also monitors ingredient levels and equipment performance. When it detects shortages or anomalies, it alerts staff before service breaks down. Microsoft positions this dual role, customer interaction and operational monitoring, as a core advantage of agentic systems over traditional automation, which typically handles a single, fixed task.
Beyond robotics, Microsoft describes a broader set of agentic store applications built on computer vision and language models. These include systems that detect empty shelves, guide shoppers to products using voice interaction, and respond to real-time changes in store traffic. Microsoft emphasizes that many of these capabilities run on existing cameras and infrastructure, reducing the need for specialized hardware.
The company says early deployments show improvements in service speed, consistency and staff efficiency, particularly during peak periods when labor shortages are most acute.
Infrastructure, Consumers Begin to Align
Microsoft’s store-level strategy aligns with findings from Stripe, which outlined three agentic commerce trends observed at NRF 2026. Stripe reported that large retailers are actively restructuring product catalogs and back-end systems so autonomous agents can reliably search, compare and transact without manual intervention.
Stripe also found that retailers are pursuing two parallel paths: building their own branded agents while simultaneously integrating with third-party agents operated by platforms and AI assistants. That dual approach is driving demand for standardized infrastructure that allows multiple agents to transact against the same merchant systems without custom integrations for each use case.
To support that shift, Stripe highlighted its work on agent-ready checkout and payments flows, including integrations that allow AI assistants to complete purchases inside conversational interfaces. Microsoft’s Copilot Checkout, which supports transactions with partners such as Stripe, Shopify and PayPal, reflects that same design principle: commerce executed inside AI-driven interactions rather than redirected web sessions.
Another emerging trend reported by Stripe is retailers introducing their own agentic commerce experiences. For example, Home Depot’s AI companion, Magic Apron, operates exclusively on the company’s website, where it delivers specialized customer support. Because it has access to Home Depot’s customer and purchase data that third-party agents do not, Magic Apron can provide more personalized assistance while building on the trust consumers already place in the brand.
Recent PYMNTS Intelligence report surveyed U.S. adults about their willingness to let AI act on their behalf across everyday tasks. It found that most consumers express at least some interest in delegating both decisions and execution to AI assistants in commerce-related activities, particularly routine or familiar ones such as reordering household goods or managing recurring transactions.
However, the study emphasizes that broader adoption will be shaped by “payments-grade trust,” meaning that transparent controls, clear data governance, and trusted authorization methods anchored in established financial credentials remain essential for consumers to feel comfortable with autonomous purchasing.
