AI commoditises general expertise
AI is making publicly codified expertise abundant; the gap between an expert and a competent AI-equipped generalist is narrowing, and that gap is where professional fees live.
Consider what it has meant, historically, to be a good accountant, lawyer or consultant. The value lay in the knowledge itself — the tax code, the precedent, the ability to diagnose an organisational problem from symptoms. That knowledge was hard-won and scarce, and the fees the professions charged were, in part, rent on that scarcity.
AI is making it abundant. Not perfect, not autonomous, but good enough that the gap between an expert and a competent person with AI access is narrowing fast. That gap is where professional fees live.
Why it happens
Public expertise rests on knowledge that is, by construction, codifiable — tax codes, accounting standards, case law, published strategy frameworks, methodology textbooks. Codifiable knowledge is precisely what AI reproduces best. The more a professional’s value depends on knowledge that exists somewhere a model could have been trained on, the more exposed that value is to commoditisation.
The exposure is not confined to a single profession. Compliance revenue is structurally threatened is the most visible case because compliance work is unusually pure — codified knowledge, standardised outputs, volume-based pricing — but the mechanism extends to first-pass legal analysis, routine advisory memos, introductory strategy work, and the codifiable portions of engineering and accounting practice.
What it implies
Fees attach to whatever is scarcer than the codifiable layer. Two categories persist above it: privileged knowledge of a specific client’s situation, accumulated through direct engagement, and trust, which AI cannot accrue on its own. The substance of The relationship is the product follows from this — once the underlying expertise is commoditised, what remains to pay for is the relationship that gives a human access to context AI cannot obtain.
For firm strategy, the commoditising layer is not where margin can be defended. Competing on efficiency with AI-native providers — including emerging AI-as-a-labour-service offerings — is a losing proposition. The useful response is to reduce dependence on the commoditising layer and invest deliberately in the adjacent one that does not commoditise.