RECO
    Back to blog
    AI & Strategy

    AI in Corporate Decision Making: What Do Your Decisions Really Rest On?

    Mag. Paul H. SchindlerMag. Paul H. SchindlerCEO, RECOPublished Last edited 3 min read

    I'm convinced that many companies already make decisions with AI. They just don't know it. And they know even less about what the AI bases its suggestions on. The endless loop explores the value of retrievable reasoning for later discussions. Knowledge transfer helps decision makers understand the operational context behind a paper instead of receiving a polished conclusion with no account of its origins.

    That was the core of my keynote on "AI & Corporate Decision Making". My thesis there was an uncomfortable one. The biggest risk is not the big, officially approved AI decision. It sits in the invisible groundwork: summaries, prioritizations and decision papers that are already produced with AI support, on top of a knowledge base nobody has checked. For SMEs, knowledge validation checks the underlying claims, while knowledge preservation retains decisions and assumptions needed to trace a recommendation.

    The opposing view is common: AI hasn't reached the decision process yet, and that will only come with the official rollout. I think that is a mistake with consequences. If you think that way, you govern the rollout and miss what is already happening.

    Where AI already shapes decisions

    My claim: hardly any decision in a mid-sized or large company is made without AI influence anymore. Rarely directly, but indirectly all the time:

    • in summaries and meeting notes
    • in research and data interpretation
    • in prioritizations and decision papers
    • in recommended actions

    Whoever writes the briefing paper shapes the decision. What's new is that often nobody can say where a particular sentence in it came from.

    A typical pattern: before a quarterly review, a division head has AI summarize the notes from her customer meetings. Some of those meetings were never written up because they happened on the phone at short notice. The summary still reads smoothly and convincingly. It just leaves out precisely the customers who threatened to cancel. The AI did nothing wrong. It simply never had the information.

    Five questions for every AI-assisted decision paper

    That makes "Should we use AI?" an outdated question. I recommend checking every AI-assisted paper against five questions.

    1. Which data went into it?
    2. Which information is missing, and why?
    3. Which perspectives are over-represented?
    4. Which are under-represented or missing entirely?
    5. Who last changed the underlying material, and why?

    My rule: if nobody in the room can answer the first and the fifth question, the paper goes back, however good it sounds.

    The real problem sits underneath the AI

    My assessment: in many corporates and larger SMEs, these questions fail at the same point. There is no clean knowledge base. Knowledge lives in people's heads, in scattered meetings, in PDFs, in chat threads, in gut feelings and in decisions whose reasoning nobody remembers. Then we wonder why the AI sometimes makes "strange" recommendations. Garbage in, garbage out. That holds for AI as for any spreadsheet analysis.

    A stronger model won't fix this, it only phrases an incomplete foundation more persuasively.

    What we take from this at RECO

    That is why we work so persistently at RECO on knowledge documentation that comes out of everyday work, with no extra effort for people who already sit in too many meetings. Documentation isn't glamorous. But good decisions are only as good as the knowledge they stand on.

    RECO captures conversations such as jour fixes, handovers and decision meetings, makes their content searchable and records how it changes over time. Every recommendation derived from it shows which conversation it comes from. People decide, AI supplies the context. The principle is explained in our post on human-in-the-loop as a core principle.

    This points to the use cases where I see the biggest lever in corporate settings:

    • onboarding across multiple locations
    • employee development with a documented learning curve
    • knowledge transfer ahead of upcoming retirement waves
    • handovers during leadership changes and after M&A transactions
    • leadership decisions with a fact base you can trace
    • faster, auditable business processes

    My forecast

    My assessment: soon nobody will ask whether a company uses AI. Everyone will. The difference will be who has a well-kept, living knowledge base in which the AI finds something solid. Everyone else will get very persuasively written papers built on thin foundations.

    Our AI readiness checklist gives you a rough sense of where your organization stands. You can put the five questions above to the next paper that lands on your desk.

    Related articles

    In a 30 minute software demo we show you how RECO secures and transfers knowledge in your company. Request a demo now and we'll get back to you with a proposed time.