Your rivals bought the same AI you did. So what actually becomes your advantage?

You sat through the vendor demonstration, and so did your closest competitor, and it was the same demonstration. The same model, the same benchmarks, the same confident promises about productivity. Somewhere in that room a quieter, more strategic worry forms, the kind that does not fit neatly into the business case: if everyone can buy exactly this, how does any of it become an advantage that is actually yours? You can feel that a tool available to all is, by definition, an edge for none.

Here is the resolution, and it is more reassuring than the worry. The advantage was never going to be the tool. A 2025 systematic review in the Journal of Innovation & Knowledge, which synthesised 1,377 high-quality studies, lands on a single idea: AI's durable contribution is not automation, it is knowledge creation, and knowledge is the one asset a competitor cannot buy or copy. A model is a purchasable, imitable resource. What your organisation learns by using it, embedded in your routines, your people and your culture, is socially complex and continually co-created, which is exactly what makes it a moat. The bottleneck is no longer the technology. It is how well you turn the technology into knowledge.

Why is knowledge the advantage, not the AI itself?

Because a resource everyone can acquire cannot, on its own, differentiate anyone. The review builds its case on two long-standing strategy ideas: the resource-based view (which holds that durable advantage comes from resources that are valuable, rare, hard to imitate and hard to substitute) and its extension, the knowledge-based view (which treats organisational knowledge as the strategic asset that meets all four tests). An AI licence fails three of those four; any rival can buy the same one tomorrow. Knowledge passes all four, because it lives in routines and relationships, resists copying, and keeps evolving.

AI's role in this is to accelerate the knowledge, not to be it. Machine learning turns data into insight that informs decisions, and used well it amplifies human cognition rather than replacing it. That is the honest promise: not a workforce swapped for software, but an organisation that learns faster because people and AI think together. It is the same reason you cannot out-hustle the machine: the edge is not effort or even the tool, it is the quality of what you come to know.

Why do most organisations miss this?

Because they buy AI as a cost to cut rather than a capability to build. The review is candid that the literature on AI in management is fragmented, with little consensus on how to embed it in managerial functions, and that adoption is uneven: only about a third of supply-chain executives are actively using AI to transform their operations. Buying the tool is the easy part. Assimilating it, reshaping how the work flows so the organisation actually learns from it, is the hard part most skip.

This is the failure behind why most organisations fail at AI adoption, and the evidence is consistent across very different studies. PwC's 2026 performance research found that the roughly one-fifth of companies capturing around three-quarters of AI's economic value win by using it to reinvent the business, not by deploying more tools. Same finding, different words: the advantage is in what you do with AI, and what you learn from it, not in owning it.

How do I turn AI into an advantage only I have?

By building a capability, not buying a product. The review frames it through dynamic capabilities, the organisational muscles of sensing, seizing and reconfiguring. Translated into practice, it looks like this.

1. Sense. Use AI to scan for the patterns and opportunities you would otherwise miss, and treat what it surfaces as raw material for strategy rather than a report to file.
2. Seize. Act on what you sense. Build the internal capability to turn an insight into a decision and a product, quickly, while the opportunity is still open.
3. Reconfigure. Reshape how the work flows so the organisation absorbs the technology, rather than bolting it onto business as usual. Assimilation is where the learning actually happens.
4. Turn insight into shared knowledge. Capture what people learn while using AI and spread it across the firm. Knowledge that stays in one head is not a moat; knowledge held by the organisation is.
5. Govern for trust. Accountability, transparency, privacy and security are what let people rely on AI's output, and the review names them as preconditions, not afterthoughts. These are exactly the questions a board should be asking.

What does this change for me as a leader this quarter?

It reframes the AI budget from a purchase into a capability build. The question stops being "which tools do we buy" and becomes "what will our organisation know in a year that our competitors will not, because of how we used AI." That question points at learning, at reshaping the work, at governance, and at the long term rather than the quarter.

That long horizon is the review's deeper point. It ties AI's real value to countering short-termism, weak stakeholder engagement and wasted resources, and to building an advantage that is sustainable in both senses of the word. Think of it like the growth rings of a tree: the advantage is laid down slowly, ring by ring, and cannot be faked or bought in a hurry. The leaders who pull ahead this decade will be the ones who used AI to learn faster than everyone else, and let that knowledge compound. The best time to start this year's ring is now.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.