The AI Impulse plan, announced on September 22, 2026, by the Ministry of Labor and Solidarity, has repositioned vocational training around a specific challenge: the operational mastery of artificial intelligence by workers. This plan mainly reveals a structural gap between the rapid adoption of AI tools in companies and the actual level of preparedness of employees.
Regulatory obligation for AI skills development: what Article 4 of the European regulation changes
Since February 2, 2025, Article 4 of Regulation (EU) 2024/1689 on artificial intelligence imposes an explicit responsibility on suppliers and deployers of AI systems: to ensure a sufficient level of mastery of AI among those who use these systems on their behalf. This obligation does not translate into a single certification but into an adaptation of training paths according to the level, experience, and work context of each employee.
We observe that this regulatory constraint is changing the game for training managers. It is no longer about offering a generic module on AI, but about mapping actual uses by position and linking them to targeted learning. A payroll manager using an automation tool does not have the same needs as a communications officer using a content generator.
Companies deploying AI systems without documenting the associated training expose themselves to a risk of non-compliance. The regulation does not foresee a direct sanction on Article 4, but it creates a standard of diligence that internal audits and social partners can invoke. Among the training courses offered by Formaxio, some paths already integrate this compliance logic by structuring skills development according to the level of exposure to AI tools.
Revision of professional titles and certifications: integrating AI into frameworks
The AI Impulse plan aims to accelerate the updating of professional titles and educational content to incorporate the uses of artificial intelligence. Specifically, this means that certification frameworks will evolve beyond technological fields. Professions in commerce, management, logistics, or human resources are directly concerned.
This revision poses a technical problem for training organizations: how to assess an AI skill that evolves faster than the usual certification update cycles? The answer lies in short, modular formats that can be revised without questioning the entire certifying path.
We recommend that HR development managers monitor publications from France Compétences on this subject. The skill blocks related to AI will likely appear as additional modules attached to existing titles, rather than as new standalone certifications.
In-person or digital training: deciding based on the type of skill targeted
The format should not be chosen for convenience but for pedagogical suitability. A procedural skill (using an LMS, configuring an automated reporting tool) is effectively acquired through asynchronous digital learning. A relational or decision-making skill (analyzing a result produced by AI, arbitrating between contradictory algorithmic recommendations) requires in-person training or at least synchronous training with situational practice.
Blended learning is not an innovation in itself. What changes is the granularity of the decision-making. The most effective paths segment each educational objective and assign it a distinct format:
- Theoretical knowledge (AI vocabulary, regulatory framework, operating principles) is delivered through short e-learning modules, accessible on mobile
- Critical analysis skills (evaluating the reliability of an AI output, identifying a bias) require supervised workshops with real cases from the learner’s profession
- Management skills (integrating AI into an existing workflow, training a team on a new tool) are developed in cohorts with professional coaching support
This segmentation avoids the trap of a 100% digital path that generates low completion rates for complex skills and a 100% in-person path whose cost per learner becomes prohibitive at the organizational level.
Building an AI skills development path for non-technical employees
The majority of employees concerned by AI are not developers. Yet, training catalogs still overrepresent technical paths (Python, machine learning, data science) at the expense of business uses. The AI Impulse plan explicitly targets this gap by aiming to democratize access to AI skills beyond technological profiles.
A relevant path for a non-technical profile revolves around three progressive levels:
- Understanding what the AI used in their position does (general functioning, limitations, documented error cases)
- Knowing how to formulate queries or configure the tool to obtain a usable result (applied prompt engineering, filter configuration)
- Being able to evaluate the relevance of the result and decide when not to follow the algorithmic recommendation
This third level is the most neglected. It requires a judgment skill that is not developed through an online quiz but through repeated exposure to ambiguous situations. Companies investing at this level observe a more sustainable adoption of AI tools by their teams.
The updating of professional titles announced as part of the ministerial plan should facilitate the formal recognition of these skills. For organizations not waiting for this deadline, the immediate lever remains the construction of internal paths aligned with actual uses, validated by evaluations in work situations rather than by disconnected multiple-choice quizzes.
Vocational training is entering a phase where regulatory compliance and operational efficiency converge. Organizations that structure their AI skills development paths now, distinguishing levels of exposure and suitable formats, will gain a lead that will be difficult to catch up for those who settle for ticking the box of the generic e-learning module.
