The European AI Omnibus, which came into effect on July 27, 2026, has reshuffled the deck for any company integrating artificial intelligence into its processes or offerings. This new framework, combined with the rise of ESG reporting and the reshaping of value chains, requires leaders to simultaneously rethink their governance, technology stack, and skills strategy.
AI Governance in Business: What the 2026 AI Omnibus Changes
The AI Omnibus regulation, applicable since July 27, 2026, has postponed certain heavy obligations while simplifying compliance for SMEs and small mid-sized enterprises. The chosen logic favors experimentation through regulatory sandboxes accessible to mid-sized structures.
In practical terms, this means that a company deploying a customer scoring tool or HR automation must now document the risk level of its AI system, but has extended deadlines if it falls within the SME scope. We observe that many leaders still confuse GDPR compliance with AI Act compliance: both are cumulative, and the sanctions provided by the CNIL as the supervisory authority reach significant ceilings.
The most common mistake is treating AI governance as a purely legal issue. Companies that gain a competitive advantage from this framework are those that integrate compliance from the product design phase, not those that delegate it to an external firm once deployment has begun. Publishing in the SLR magazine on Rue du Business allows one to keep track of these regulatory developments applied to French SMEs.
CSRD Reporting and ESG Data: An Underestimated Business Lever
The first wave of sustainability reporting for the 2025 fiscal year affects large European companies subject to the CSRD, with increased requirements for limited assurance and structured ESRS formats. The SMEs supplying these large accounts are directly impacted, even if they are not themselves subject to it.
A client subject to the CSRD will require its partners to provide carbon data, social indicators, and supply chain traceability. Failing to provide these elements amounts to disqualifying oneself from certain tenders.
We recommend that SMEs start structuring a minimal foundation for collecting ESG data now, even informally. A well-designed spreadsheet is better than poorly fed specialized software. The goal is not to produce a complete CSRD report, but to respond to supplier questionnaires without improvisation.
Data to Prepare as a Priority
- Scope 1 and 2 carbon footprint, even estimated: the simplified methods published by environmental agencies are sufficient for an initial usable approximation
- Basic social indicators (turnover, training, workplace accidents) already available in most HR management tools
- Traceability of tier 1 suppliers: country of origin, any certifications, contractual conditions on responsible practices
Hybrid Skills and Talent Management: The Real Bottleneck
The sought-after profiles combine industry expertise with mastery of generative AI tools. A management controller capable of building structured prompts to analyze datasets is more valuable than a junior data analyst without sector knowledge.
Lifelong learning is no longer just an HR slogan. The companies that progress are those that allocate a training budget indexed not to the payroll, but to the gaps identified against their technological roadmap. Training three key individuals on a specific tool produces more impact than offering a catalog of generic training to all employees.
Structuring a Skills Development Plan
We recommend starting from the operational need, not the training offer. Identify the three most time-consuming business processes, assess which can be partially automated, and then train the relevant teams on the selected tools.
The classic trap is to multiply SaaS licenses without support. A project management tool or a CRM enhanced by AI only produces value if end users actually change their work habits. Real adoption takes precedence over technical deployment.
Customer Proximity and Data-Driven Communication: Balancing Tools and Strategy
Personalization technologies now allow for segmenting a customer base with a granularity that was inaccessible five years ago. The temptation is to stack tools: CDP, marketing automation, advanced analytics. In practice, most SMEs do not even leverage the data already present in their existing CRM.
Before investing in a new tool, we recommend an audit of the existing setup. How many fields are actually filled in the customer record? Are purchasing behavior data linked to communication data? A clean and utilized database is worth more than a poorly configured sophisticated tool.
- Clean the CRM database: remove duplicates, standardize fields, enrich records based on recent interactions
- Link communication channels (email, phone, social media) to a unified customer view before considering any automation
- Measure customer acquisition cost by channel to allocate marketing budgets based on real data, not intuition
Customer proximity is not just about technology. Service companies that maintain an identified human point of contact, with a stable interlocutor, show significantly higher retention rates than those that fully switch to automated interfaces.
The European regulatory framework, the ESG pressure from clients, and the reshaping of skills converge towards a common requirement: companies that structure their governance, data, and teams around these three axes gain resilience. Those that treat these issues separately, or postpone them, accumulate organizational debt that ultimately impacts revenue.
