AI, Work, and the Skills of Tomorrow

Training employees on Artificial Intelligence has become a priority for many organizations.

But what kinds of learning truly matter?

Recent studies from Stanford and the OECD suggest that before expanding training catalogs, we should first understand how AI is transforming work itself.

Technology Is Advancing Faster Than Our Reference Points

AI use is advancing faster than our understanding

Stanford’s AI Index 2026 report indicates that a large majority of the organisations surveyed say they use AI and generative AI in at least one function.

These figures do not represent all French businesses, but they do point to widespread use among the organisations studied.

The OECD also notes the development of AI systems that can carry out sequences of actions in pursuit of a goal.

This research invites us to shift our focus.

When a tool takes over part of an activity, training needs depend on what the person still needs to understand, decide and oversee.

Valuable Skills Go Beyond Tool Mastery

Useful skills go beyond mastering a tool

Training needs are not limited to programming. Using, analysing and interpreting data matter just as much as:

• problem-solving;

• critical thinking;

• management skills;

• the ability to explain a decision.

Relevant training must always be connected to a real workplace situation.

Experience using AI matters too

Recent research shows that experienced users often tackle more complex tasks.

This suggests an important approach for learning and development leaders:

• allow time for experimentation;

• encourage people to share what they learn from experience;

• support the use of AI in day-to-day work.

A convincing demonstration is not enough to establish a lasting professional practice

Start with the work to build the training plan

Our proposal is simple:

• identify situations where AI is already being used;

• understand which tasks are changing;

• identify the new checks that are needed.

Assessment can then focus on observable abilities:

• selecting the right information;

• checking a result;

• explaining why a tool is being used;

• recognising the limitations of AI.

As technologies evolve, we need guidance that evolves with the realities of work.