Agentic AI Will Reshape Commerce, But Only Infrastructure Can Unlock It

by editorial on 29th Sep 2025 in News

As AI assistants evolve into agents capable of real transactions, the ad industry faces a new challenge. The missing piece is not another model but the infrastructure that connects merchants, enables attribution, and ensures monetisation. Aaron Simpson, founder of Kindred, explains why this invisible layer is about to become the most strategic asset in advertising and commerce.

"Agentic AI" has quickly moved from theory to boardroom conversation. This is not just generative chat tools getting flashier. Agentic systems can take action on behalf of users, making decisions, executing transactions, and learning from feedback. That shift fundamentally changes the commercial stakes.

For years, the conversation around AI revolved around capability. Could a model write copy? Could it summarise a document? Could it answer questions in a way that felt human? Those questions now feel dated. The horizon has moved. Today the real question is whether AI can move from passive to active, from information retrieval to transaction execution. And if it can, who will provide the plumbing that makes this commercially viable?

At present, the answer is nobody at scale. Platforms are beginning to embed transactions into AI-native experiences. Consumers are already testing what it feels like to buy through a conversational prompt. Yet the supporting infrastructure is missing. Agents need real-time access to pricing, inventory, merchant catalogues, and attribution systems. They also need to understand which monetisation paths to prioritise. Without this layer, agentic AI will remain a promising demo rather than a functioning market.

The gap is clear. Huge sums are being poured into models and compute, but merchant connectivity remains fragmented. Most brands and retailers do not have feeds that translate cleanly into the decisioning logic of agents. The existing attribution systems are still designed for media-first ecosystems, not autonomous decision engines. This disconnect is why so many AI-powered shopping demos still feel like prototypes rather than true commerce platforms.

The opportunity lies in the invisible layer between agents and merchants. This infrastructure must translate offers into AI-ready formats, enforce rules and brand policies, and trace outcomes with accuracy. It is not enough to recommend a product. The system must also record the value exchange, ensure the transaction is attributable, and allow the brand to see how spend links to performance. In short, the infrastructure has to rebuild the foundations of digital advertising for a new environment where the "ad slot" is a natural language request.

Retail media has already shown how quickly spend can shift when data and purchase signals are unified. Agentic AI takes that logic further, collapsing the funnel so that discovery, consideration and purchase may happen inside a single conversation. That rewrites the economics of advertising. Agencies will need to plan not around impressions but around decision influence. Brands will compete not just on creative but on how well their commercial logic is integrated into agent ecosystems.

At Kindred, we have been focused on this challenge directly. Our work has centred on wiring fragmented merchant data into AI-native formats and building the rails for attribution, decision logic, and commercial viability. The aim is not to launch another assistant. It is to ensure that every assistant has the confidence and capability to transact in ways that respect both consumer intent and brand value.

If 2025 is the year agentic AI moves from hype to scale, the winners will not be defined by who builds the most impressive model. They will be defined by who controls the infrastructure that allows transactions to flow reliably, transparently, and at scale. History shows that infrastructure quietly wins twice: once when adoption explodes, and again when it becomes impossible to remove.

The race for this layer has already started. Those who invest now in building it will shape the future of AI commerce for the decade to come.