Sharper, Not Smaller: The Business Analyst in the Age of AI

Day 1 | The Day I Stopped Fearing AI and Started Directing It

I will be honest with you. For a stretch over the last few years, I felt the ground move under my feet. Every week brought a new headline. AI writing requirements. AI generating user stories. AI summarising stakeholder calls. AI building process maps from a single paragraph. I had spent years getting good at those things, and now a tool could produce a rough version in seconds. If you have felt a quiet version of that worry, you are not alone, and you are not wrong to take it seriously. I want to be clear that the worry is not silly. The tools really are good at the visible outputs of our work, and visible outputs are what most people, including a lot of managers, mistake for the whole job. If your sense of your own value is tied to producing those artefacts, watching a machine produce them in seconds is genuinely unsettling. I felt it. The way through is not to deny the feeling. It is to look harder at what your value actually was, and to find, with some relief, that it was never the part the machine can do.

Then something happened in a meeting that changed how I see all of it. A junior analyst on a project I was advising had been asked to pull the requirements together for a new claims process. He was sharp and eager. He fed the discovery notes into an AI tool, asked for a full requirements document, and it gave him something that looked the part. Clean structure. Confident language. Proper headings. He barely touched it. He walked into the review and presented it as his work. The sponsor read for about a minute, looked up, and asked a simple question. Why have we excluded partial payments from scope? Silence. The document had a line about it. The AI had written that line. But nobody in the room could explain the thinking behind it, because there was no thinking. The tool had pattern-matched to what claims documents usually say. It had not sat in the workshops. It had not heard the operations lead mention, almost in passing, that partial payments were the single biggest source of customer complaints. That context never reached the notes, so it never reached the output.

The meeting stalled. Trust took a knock. And the analyst learned, in the most public way possible, where the real line sits between what AI can do and what only he could do.

The fear was real, but it was pointed at the wrong thing

That moment turned my fear into something far more useful. Clarity. The AI did not fail in that room because it was weak. It failed because it was handed something it can never hold: judgment grounded in context, defended in front of people who matter. The mechanical part of the work- the drafting and structuring and summarising- it did in seconds. The part that made the analyst valuable, knowing why a decision was made and being able to stand behind it, was never the AI's to carry. So the question stopped being Will AI replace me. That is the wrong question, and it pushes you into one of two bad responses. You either panic, or you pretend nothing is changing. Both leave you worse off.

The better question is this. Where does my judgment add the most value, and how do I clear everything else off my plate so I can spend more of my week there?

I have watched both wrong responses play out. The panicked analyst treats every new tool as a threat, refuses to touch it, and slowly becomes the person on the team who cannot keep up. The one in denial does the opposite. They paste AI output into deliverables without reading it, tell themselves nothing has really changed, and sooner or later have their own version of the partial payments meeting. Neither is thinking about where their value sits. They are both reacting to a headline instead of looking at the work.

That one reframe is what this whole series is about.

The principle underneath all of it

Here is the idea I want to leave with you today, because everything else this week sits on top of it. AI is very good at producing the artefact. It cannot own the accountability. A requirements document is an artefact. The decision about what is in scope and why, defended in a room full of competing interests, is accountability. A meeting summary is an artefact. The read on which stakeholder went quiet for a reason and needs a private follow-up is accountability. A process map is an artefact. The judgment that one of those steps exists only because of an old workaround nobody wants to admit to is accountability. The analyst who blurs the two, who hands over the accountability along with the artefact, will struggle. Not because the tool is bad, but because they have given away the one thing that made them necessary.

The distinction sounds abstract until it costs you something. Then it becomes the most practical idea in your toolkit. Every time you are about to hand work to a tool, you can ask one question. Am I handing over the artefact, or the accountability for it? The first is fine and often smart. The second is how careers quietly stall. The analyst who keeps the accountability and lets the tool handle the artefact gets something genuinely useful. More time. More range. More room to do the thinking that actually moves a project forward. Let me put it another way, because this is the hinge the whole week turns on. Imagine two analysts handed the same workshop notes. The first spends the afternoon typing a tidy requirements document by hand and walks in proud of the effort. The second has a tool draft that same document in minutes, then spends the afternoon chasing the three contradictions buried in the notes, having a quiet word with the stakeholder who went silent, and confirming why one requirement that looks redundant is actually holding the whole thing up. Both produce a document. Only one produced understanding. When the hard question comes in the review, only one of them can answer it.

Where the week is going

Over the next four days, I will walk you through the framework I now use every day.

Tomorrow, what is safe and smart to hand to AI, the work that quietly drains your hours without using your real skill. Wednesday, what you should never hand over, the parts of the role that are the reason you get hired and trusted. Thursday, how to build AI into your actual workflow, step by step, in a way that holds up under scrutiny. Friday, why all of this makes the analyst role sharper rather than smaller, and how to position yourself so the people around you see it too. That junior analyst, by the way, is doing well now. The meeting that embarrassed him became the lesson that sharpened him. He still uses AI every single day. He just never lets it walk into a room on its own again.

For today, sit with one distinction. Artefact versus accountability. Once you can tell them apart, the fear starts to fade, and something steadier takes its place.

Go out and be successful.

Oluwatosin Ogunkoya | Flotog BA Insights | www.flotogbainsights.com Tomorrow: What to Hand to AI. The safe list of work you can delegate today, the tasks that quietly drain your hours without using your real skill, and the one question that tells you in seconds whether something belongs on it.