Target's Senior Vice President Siobhán Mc Feeney rejected the notion that AI models themselves create competitive advantage. Speaking at VB Transform 2026, she argued that the infrastructure, processes, and systems built around models define the real moat.
"The models are great, and they're important. They're just not sufficient to be the competitive advantage," Mc Feeney said. She emphasized that Target's edge comes from everything layered on top of the foundation model, not the foundation itself.
This perspective challenges the prevailing industry assumption that owning cutting-edge AI models drives business differentiation. Mc Feeney contended that most enterprises chase AI agents indiscriminately. Not every business problem requires an agent, she argued. The retail giant applies strict discipline to agent deployment, requiring them to earn autonomy over time rather than granting it by default.
The stance reflects a maturing approach to enterprise AI adoption. As large language models become increasingly commoditized, with capable options available from OpenAI, Anthropic, Google, and others, the actual value accrues to companies that can architect effective implementations. Data pipelines, prompt engineering, retrieval-augmented generation systems, monitoring frameworks, and integration with existing business logic matter far more than which model runs underneath.
Target's philosophy mirrors how software companies think about infrastructure. AWS didn't dominate cloud computing because it built the most innovative servers. It dominated because it wrapped those servers in APIs, tooling, and operational discipline that became too valuable to replicate.
For retail specifically, this means AI agents need to connect seamlessly with inventory systems, pricing engines, customer data, and fulfillment networks. A model running in isolation accomplishes nothing. The orchestration layer, the decision-making rules, the feedback loops that refine behavior—those are where competitive advantage lives.
Mc Feeney's comments