Imagine LinkedIn had existed in 1979. And someone had posted: GRUNDIG will go bust before long, its competitive advantage will soon be zero.
How would you have reacted? I recently read up on the company’s history again. It’s incredible, and it genuinely shook me. In 1979 I would have said: nonsense. A show-off.
The CFO function, like other functions in a company, is exactly this kind of thing that feels impossible.
These documented cases show where things are heading:
• The Big Four are investing billions in automating their own core business
• JP Morgan has rolled out its “LLM Suite” to more than 140,000 employees
• Amazon says it is cutting its white-collar workforce substantially, including in finance
Try searching for Dukaan e-commerce or Blue Focus marketing together with “employees replaced by AI”.
There isn’t room to go into more detail here, but ChatGPT and the like will happily fill in the gaps.
Of course, the opening line is an exaggeration. And before anyone points it out: Klarna has already rowed back.
There would also be serious downsides and risks:
• The four-eyes principle goes out of the window
• Audit and liability risks pile up
• False precision
• Cognitive bottlenecks, absurd workloads and concentration risk
• Huge dependence on individuals
• The talent pipeline dries up
But:
• It’s already possible to say, function by function, how much can be automated: accounting, controlling, financing, M&A and so on (in descending order, incidentally).
And companies with legacy systems built up over many years will find it harder than start-ups.
• With data consolidation and forecasts available in real time, the role will shift even further towards making decisions between scenarios.
• Decisions on data architecture will no longer be made in IT; the CFO needs to carry the blueprint in their head.
• Orchestration will become a top priority: which systems and which outputs can you trust, and at what point does a human take over?
• Data hygiene and interfaces will matter enormously and will become a board-level issue overnight.
• Spotting hallucinations and bias will become an important skill.
• Relationships with all stakeholders will matter more than ever, wherever the line between human and machine falls, and that line is moving all the time right now.
• All of this is already changing our own work in executive search, in how we select and interview.
In the end, you have to trust your AI and stand by it. That isn’t always easy, especially when things are uncertain.
Note: the original German post was not written by AI; this English version is a translation. The idea comes from interviews we conduct with investors, CEOs, CFOs and other stakeholders.
