Industry news

AI cannot replace human business understanding

Published May 2, 2026

There’s no doubt AI has changed the rules of marketing. It’s a real productivity shift that continues to get better all the time. But as the technology matures, its limitations also become clearer. Brand positioning is one of the areas where AI cannot yet replace human work.

What AI sees when the tool looks at your brand

AI tools like ChatGPT, Perplexity, and Google’s AI overviews continuously collect information about companies online. They read claims, seek external confirmation, and assemble a consolidated picture. What they do well is connect facts and draw logical conclusions. If A is true and B is true, then C is likely true. That is the obvious conclusion.

SEO expert Jason Barnard describes this gap with the concept of “the framing gap” in an article series on AI visibility published on Search Engine Land. His point is that AI can pass on information about a brand, but only the version that logically follows from what can be substantiated. Choosing the most strategic version is still a human decision.

What is commercially interesting is rarely C. Barnard instead calls it J (A + B = J), meaning a conclusion that requires a business purpose to even be chosen. AI moves to C on its own, but J requires someone with business understanding to choose that specific conclusion and explicitly state the connection. Then, and only then, can AI carry it forward.

The conclusion AI draws on its own therefore rarely benefits the brand commercially. To reach J, someone who understands the company, the market, and the business needs to point out the conclusion and build the logical bridge to it. Behind it lies an unpredictable, strategic choice with business intent — and that is something AI fundamentally lacks.

Three levels of AI visibility

Barnard outlines three levels of how a brand can be visible in AI answers:

Level 1: Scattered evidence. The company makes claims on its website, but the connection to concrete evidence is missing. AI has to guess, which often results in an uncertain or absent answer.
Level 2: Linked evidence. Claims are explicitly connected to evidence through text, links, and structured data. AI can pass them on with confidence. Already at this level, a niche company with well-linked evidence can outperform a larger company with scattered information.
Level 3: Framed content. At this level, AI passes on exactly the narrative the brand has chosen. This requires the communication to contain more than evidence — the brand also explains what the evidence points to and why it matters.

It is level three that requires strategic human input. And that is precisely what separates the brands AI chooses to recommend from those mentioned superficially or not at all.

AI amplifies the strategy you already have

It would be wrong to read this as criticism of AI tools. They solve real problems and they do it well — analysis, content production, advertising, tracking, and automation of the repetitive. We at BBO work with them daily. But the tools can really only amplify a strategy; they cannot build one or come up with something new.

If you have already done the work of explaining what your brand stands for, why it is needed, and what conclusion AI should draw, a better AI model will take that narrative and convey it even more clearly and persuasively. But if you have not put in that work, a more advanced model will not help you further. It still has no stable foundation to build on; it just produces a more well-articulated answer without strategic direction.

That is precisely why we need to view brand work and content strategy as communication to machines as much as to people. It is not enough to be clear to a reader; the content needs to give AI enough context to assemble and convey the right picture. That work requires a deep understanding of your market, your customers, and your business strategy. AI can structure and distribute, but the starting point must come from you.

The gap grows as AI gets better

Barnard calls it a positioning gap (framing gap) and argues that it widens as the technology develops. Those who give AI systems what they need — clear claims, linked evidence, and a well-chosen interpretive framework — gain more visibility and more credibility. Those who do not, see an ever-larger share of potential AI visibility go to competitors.

Many marketers will recognize this from SEO: those who make it easier for search engines win the space. Now the same logic applies to GEO and AI visibility, and the stake is what a potential customer sees when they ask an AI tool for a recommendation.

Frequently asked questions

How do we know what level we are at?+

Search for your company in ChatGPT or Perplexity and read the answer carefully. Is it brief or uncertain? That suggests the evidence is scattered. Is it correct but flat, without explaining why you are relevant? Then you are probably at level two. If AI passes on your narrative with context and clear connections, you are likely at level three.

Does this mean we should reduce the use of AI in content work?+

On the contrary. AI is a powerful tool for structuring, scaling, and distributing content. What is required is that there is a well-thought-out positioning to amplify — that is, a narrative that a person with business understanding has chosen and built. Without that, AI spreads generic information.

What is the first step?+

Start by taking stock of your most important claims and ask yourselves whether the evidence is actually visible and connected. Then you need to decide which interpretation of your narrative you want AI systems to pass on, and write content that builds that bridge.