AI and training: what finance leaders need to understand
Finance leaders treat training as a cost to be contained. AI in training changes this frame: it makes measurable an investment that was not. An AI system produces data that allow thinking in terms of return on investment: reduced training time by eliminating already-mastered content (documented gains of 20% to 40% on comparable scopes), fewer retakes, traceability that secures OPCO (French vocational training funding body) funding, and measurement of skills progression against operational indicators.
Why the CFO resists training
Training is seen as discretionary spend for three structural reasons. Its return is not measured: accounting records expenditure but not yield. Its impact is delayed: effects appear at 3, 6 or 12 months, outside usual reporting cycles. Its link to results is hard to isolate without a control group.
What AI changes in the equation
An AI system changes three variables.
It reduces training time through personalisation. A learner who only covers content matching their actual gaps spends less time in training. On documented scopes, this reduction is 20% to 40%.
It reduces retakes. A system that measures actual acquisition and adapts revision reduces the rate of re-training on insufficiently acquired topics.
It produces traceability that secures OPCO funding. Funding applications rejected for insufficient documentation represent a real loss. OPCO (French vocational training funding body) files that are fully traceable skill by skill are more robust under audit.
It makes skills progression measurable against operational indicators. For the first time, it becomes possible to correlate progress on a skill with an operational performance indicator.
How to build the financial case
The most convincing method for a CFO is to start from a limited, documented, comparable scope. Identify a precise scope (one team, one or two priority skills). Measure current training costs on this scope: pedagogical cost, salary cost of training hours, cost of retakes, cost of non-compliance. Project savings: 20% reduction in training time (low, prudent threshold), 50% reduction in retakes, OPCO funding gains. Compare to system cost over 3 years.
What to avoid
Comparing platform cost to the overall training budget without quantifying the problem it solves. Presenting theoretical gains without a baseline. Omitting integration and configuration costs.