The European AI Act has changed the game for any company deploying an artificial intelligence system. As of February 2025, employers must ensure that their employees are upskilled on these tools. Business trends in 2024-2025 are no longer just about adopting the latest technology: they require navigating a tightening regulatory framework and operational trade-offs that most market analyses overlook.
AI Literacy Requirement: What the AI Act Imposes on Companies
Article 4 of the European regulation on artificial intelligence has created an unprecedented obligation. Any employer providing an AI system (chatbot, AI functions integrated into business software, automated recruitment tool) must take concrete measures to develop AI proficiency among the individuals operating these systems.
The Digital Omnibus Regulation 2026/1744, which came into effect on July 27, 2026, has relaxed the requirements: the company no longer has to prove a “sufficient level” for each employee but must demonstrate that it has implemented training and support measures. This distinction is significant from a legal standpoint.
Since August 2, 2026, this obligation is no longer theoretical. In France, the CNIL, DGCCRF, Arcom, and DGE have enforcement and sanctioning powers. We observe that the majority of SMEs have not yet formalized an AI training plan, which creates a short-term risk of non-compliance. Following business news on Actu en vrac helps identify these regulatory developments before they become operational emergencies.
High-Risk AI Systems: Classification and Risk Management in Business
The AI Act distinguishes several risk levels. Systems classified as “high risk” directly concern HR functions, credit scoring, biometric surveillance, and application sorting. An AI used to filter CVs is considered high risk and subject to enhanced obligations: technical documentation, compliance assessment, and ongoing human oversight.
This classification requires legal departments and IT departments to map all AI tools deployed within the company. A talent management software that integrates a predictive analysis module can shift into the high-risk category without the buyer being aware at the time of deployment.
We recommend a structured internal audit based on three criteria:
- Does the tool make or influence a decision affecting an individual (recruitment, rating, access to a service)?
- Has the provider supplied the compliance documentation required by Annex III of the regulation?
- Is there a human oversight mechanism in place, with documented reporting procedures?
Ignoring this mapping exposes the company to sanctions, as well as a reputational risk that markets quickly penalize.
Social Dialogue and AI: The Role of the CSE in Technological Deployment
The deployment of AI tools in a company is not solely the management’s decision. The social and economic committee (CSE) must be consulted whenever a technological project alters working conditions. AI is now a mandatory subject of social dialogue in any company with a CSE.
The European directive on work via platforms, adopted at the end of 2024, further strengthens this framework by regulating algorithmic management. Workers must be informed when an algorithm is involved in task allocation, performance evaluation, or salary determination. This directive primarily targets digital platforms, but its principles are gradually influencing labor law applicable to all companies.
In practical terms, an employer deploying an AI-driven planning tool without prior information to the CSE exposes themselves to obstruction charges. Employee representatives are becoming more knowledgeable on these topics and are making increasingly precise requests for expertise.
Anticipate Rather Than Suffer Consultations
Companies that involve the CSE from the project scoping phase of AI gain time. A co-constructed specification reduces recourse and accelerates deployment. Conversely, consulting the CSE after the production of a tool generates costly blockages.
Online Presence of TPE-PMEs: A Structural Lag in the French Market
Major technological trends mask a reality: a significant portion of French TPE-PMEs still lack a structured online presence. No updated website, no complete Google Business Profile, no content strategy. This lag directly impacts their ability to capture markets in an economy where purchasing journeys begin online, including in B2B.
Managing this digital presence becomes a competitive issue as concrete as cash flow management. Companies that invest in a minimal digital foundation (performant website, local SEO, customer reviews) capture a disproportionate market share compared to competitors that are invisible online.
- A responsive website with indexed service pages generates quote requests without active prospecting
- An optimized Google Business Profile with recent reviews improves local conversion
- Regular content on business issues positions the company as a sector reference
This is not a question of marketing budget. It is a matter of commercial survival for structures operating in local or niche markets in France and Europe.
The year 2024 will mark the transition of AI from an innovation topic to an issue of regulatory compliance. Companies that have not yet mapped their systems, trained their teams, and structured social dialogue on these issues are accumulating a backlog that will result in sanctions and loss of competitiveness. The framework is set, authorities are equipped, and grace periods are shrinking.
