Mentivis OS OpenMentivis OS EntrepriseMentivis OS Edu
Contact usDemo
All articles
M · AI & TrainingGeneral

Published 9 September 2026 · Updated 10 October 2026 · By Marie Castelli, Partner, Mentivis

What learners expect from an AI learning platform in 2026

In 2026, learners already use generative AI daily: 48% of French people aged 12 and over and 76% of managers and professionals in 2025, according to France's Digital Barometer. Public surveys show that they expect a learning platform to offer a clear link with their career progression, a path adapted to their needs, formats that fit the time they have, and genuine training in using AI, which most of them have not yet received.

Pillar: Adaptive learning and personalised learning paths: the guide

Where do learners stand in 2026?

Generative AI use has become common. According to France's Digital Barometer (Baromètre du numérique), carried out by the research centre Crédoc for the regulators Arcep and Arcom, the General Economic Council (CGE) and the National Agency for Territorial Cohesion (ANCT) with 4,145 people in June 2025, 48% of French people aged 12 and over use generative AI, up from 33% in 2024 and 20% in 2023. The share reaches 85% among 18-24 year-olds and 76% among managers and professionals (cadres). Among users, 27% use it every day, and 44% use it for "help with homework and learning".

Widespread professional use, with little support. The Training and Employment Barometer by Centre Inffo and the CSA polling institute (1,621 working people surveyed in February 2025) indicates that 68% of working people in France use AI, in a personal or professional context, and that about half use it at work. Among these professional users, 72% feel the need for AI training, but only 38% have already taken a course on the subject.

A similar picture internationally. Microsoft and LinkedIn's 2024 Work Trend Index, based on 31,000 knowledge workers in 31 countries including France (February-March 2024), reports that 75% of them use generative AI at work, that 78% of users bring their own tools, and that only 39% of users have received AI training from their company.

What motivates learners to train?

Their career progression. In the survey for LinkedIn's Workplace Learning Report 2024 (1,063 learners and 1,636 learning professionals in several countries including France), the top reason given for spending more time learning is making progress towards career goals. Next come the need to stay up to date, then training personalised to one's interests and goals. The 2025 edition of the report confirms that career progress is people's top motivation to learn.

AI in their job. According to the same 2024 report, four in five people want to learn more about using AI in their profession. In France, the AI skills working people want to develop are specific (Centre Inffo and CSA, 2025):

Desired AI skill Share of professional AI users who feel a training need
Choosing AI tools suited to one's needs 52%
Automating repetitive tasks 46%
Assessing the reliability of information provided by AI 45%
Writing effective prompts 42%
Understanding how humans and AI complement each other 38%
Managing the confidentiality of data sent to an AI 37%

What keeps them away from training?

In the 2025 Centre Inffo and CSA barometer, 47% of working people plan to take a training course within 12 months. Those who do not intend to cite first a lack of interest (29%), a lack of time (25%) and the absence of training matching their needs or expectations (20%), ahead of budget (18%).

Format matters too. A study by Guo, Kim and Rubin (2014) covering 6.9 million video-watching sessions on the edX platform shows that median engagement time does not exceed 6 minutes, whatever the length of the video.

What learners concretely expect from a platform

The data above points to six expectations:

  1. A visible link with their job and career. This is the top motivation measured by LinkedIn.
  2. A path that takes account of what they already know. Content that does not match their needs is among the main reasons for giving up. See Personalised learning paths.
  3. Short, active sequences. Time is scarce, and attention to a video drops off after a few minutes.
  4. Help available when the question comes up. Learners are used to asking an AI; a platform that makes them wait for an answer feels behind. This help should make them think rather than answer for them, as shown by research on learning with AI.
  5. Reliable information. According to the 2026 Digital Barometer, 64% of generative AI users often or always check the information provided; assessing that reliability is one of the skills they want to acquire.
  6. Training in AI itself, adapted to their job. See Which AI skills to train first.

What this does not imply

Learners' expectations do not on their own define what they need to learn. A pleasant platform with poorly designed content produces satisfied learners who are not very competent. Engagement (logins, completion) is measured separately from learning, as explained in My LMS is useless: why completion rates collapse.

Frequently asked questions

What share of French people use generative AI in 2026?

According to the 2026 Digital Barometer (Crédoc, June 2025 survey), 48% of French people aged 12 and over use generative AI, including 85% of 18-24 year-olds and 76% of managers and professionals.

Are employees who use AI trained?

A minority are. According to Centre Inffo and CSA (2025), 38% of working people in France who use AI at work have already taken a course on it; according to Microsoft and LinkedIn (2024), 39% of users worldwide have been trained by their company.

Why do employees decide not to train?

According to the 2025 Centre Inffo and CSA barometer, mainly through lack of interest (29%), lack of time (25%) and lack of training matching their needs (20%).

Should learners be consulted before choosing a platform?

Yes. Having a small group of representative learners test the tool reveals usability obstacles that decision-makers do not see on their own.

Sources

Further reading

Related products: MentivisOS Entreprise, MentivisOS Open

Learning with AI: what the research really says
What AI cannot do in training