Does AI make management consulting obsolete? Only the knowledge kind

Does AI make management consulting obsolete? The part that only sells knowledge, yes. Consulting that helps implement and stands behind the result becomes more valuable instead, says Felix Degeler, founder of Degeler Consulting (DC) in Vienna, AI consultant and project manager for mid-sized companies.

I'm a consultant, and I'll say it plainly: our industry has a credibility problem. Consultants sell advice but rarely carry the risk when it doesn't work. Deliver the slides, explain the strategy, and they're gone before anyone finds out whether the whole thing holds up.

Why the old consulting deal was fair for a long time

The deal was simple: the consultant knew more than you, and that's exactly what you paid for. As long as expertise was scarce and expensive to obtain, it earned its price.

That extra knowledge is free now. You type your question into an AI and get the same answer in seconds that someone would otherwise charge 300 euros an hour for. The knowledge advantage the classic consulting model rests on is gone.

What is knowledge consulting, and why is it losing to AI?

Knowledge consulting is consulting whose result is advice: an analysis, a concept or a recommendation that your team then has to implement itself. That is exactly the product AI now generates in seconds.

Market overviews, methods, common frameworks, best practices and checklists are publicly documented. Language models have absorbed this knowledge and phrase it to fit your question. What used to take days of research is on the table after one good prompt.

Try this: put the core question of a consulting offer to an AI first. If the answer already covers most of the proposed concept, you know how much of it is interchangeable.

What AI doesn't bring is your context and the consequences. It only knows your people, your processes and your legacy issues if you explain them. And it carries none of the cost when its suggestion fails in practice.

What remains when knowledge costs nothing?

The one thing AI can't take off your hands: someone who stays and gets to work. Not just saying how it's done, but pitching in until it runs. And standing up for it when it goes wrong.

That part used to be the exception, a nice extra for especially good clients. Today it's the only thing still worth paying for. Consulting is shifting from selling knowledge to selling responsibility.

Why clients are really buying shared responsibility rather than know-how is something we've described in more detail elsewhere.

What is the future of consulting with AI?

Consulting has a future if it sells implementation and responsibility instead of knowledge. The craft is shifting from having knowledge to asking the right questions, building context and checking results.

Good consultants use AI themselves, and openly. At DC we work with Claude ourselves and use it to build our own tools.

Responsibility stays with people. An AI can draft a plan in minutes, but someone has to decide whether it fits this company and stand behind that decision. Why we treat this as a fixed rule is explained in our article on human in the loop.

How do you recognise good consulting in the age of AI?

You recognise good consulting by the fact that it delivers more than an AI would on the same question: implementation, context and shared risk. Use these six questions to assess any offer before you sign.

First, who is responsible for implementation? If the answer is „you“, you're buying a document, not a change.

Second, what is on the table at the end? A running process, an introduced tool or a trained team is something different from a presentation that ends up in a folder.

Third, how much of this could an AI deliver? Ask the core question yourself first. Only pay for the part that goes beyond it.

Fourth, what non-public knowledge does the consultancy bring? Experience from comparable projects, access to specialists or knowledge of your industry can't be pulled out of a language model.

Fifth, how is success measured, and what happens if it doesn't come? A consultancy that won't name a result won't take responsibility for one either.

Sixth, does the capability stay with you afterwards? Good consulting makes your team more independent, bad consulting makes it dependent on the next engagement.

When pure knowledge consulting still makes sense

Pure knowledge consulting is still worth it where the knowledge isn't public or where a mistake gets expensive. In those cases you pay for access and assurance, not for information any AI already knows.

Specialist knowledge is the first case. Someone who has worked in a niche for years knows connections, contacts and experiences that never made it into a document. An AI can only repeat what has been written down somewhere.

Regulation is the second case. With data protection, the EU AI Act, employment or tax law, the current interpretation is what counts, and AI answers can be outdated or simply wrong. Here you need specialists who are liable for their advice, such as lawyers or tax advisors.

The independent second opinion is the third case. Anyone who wants to safeguard a major decision benefits from someone who scrutinises the plan and has no stake in the outcome.

Even then: buy this knowledge in a targeted, clearly scoped way, as an expert opinion, a review or a workshop. Packaged as a months-long project without implementation, you pay the price of responsibility for knowledge.

How we turned this into our working model

Getting hands-on has always been my thing. I've never built a single presentation as a consultant, and that's not an accident but a principle. Why we don't deliver slide decks comes down to exactly this point: a slide deck ends up in a folder, an implemented project ends up in a result.

For companies this means: with every consulting offer, check who is responsible for implementation in the end. If the answer is „you“, you're buying knowledge you could also get from an AI. How we see projects through to the result is described in our services.