AI has made its way into the daily management of many SMEs — content generation, accounting automation, chatbots, reporting tools. The European AI Act now governs this use, with obligations that vary significantly depending on the level of risk.
A regulation based on risk level
The AI Act doesn't treat every use of AI the same way. It distinguishes several categories:
- Unacceptable risk: uses that are simply banned outright (mass social scoring, for example) — out of reach for the vast majority of SMEs.
- High risk: systems used in sensitive contexts (recruitment, creditworthiness assessment, HR), subject to strict transparency and documentation requirements.
- Limited risk: tools like chatbots or content generators, subject mainly to a transparency obligation — informing the user they're interacting with AI.
- Minimal risk: the vast majority of everyday SME uses (writing assistants, internal automation tools), without any significant specific obligation.
What actually concerns most SMEs
For a typical SME using AI as a support tool (writing, customer service, reporting), the main issue remains transparency: clearly informing when content or an interaction is AI-generated, and documenting at minimum the tools used in sensitive decision-making processes (recruitment, staff evaluation).
The real risk isn't so much an abstract penalty as the loss of customer trust if AI use isn't clearly acknowledged — a chatbot pretending to be human, for instance, is as much a transparency problem as a compliance one.
The link to your management and accounting
Automation tools built into an ERP like Odoo (invoice recognition, accounting categorisation suggestions) generally fall under minimal risk — but vigilance remains necessary as soon as AI is involved in a decision that affects a person (HR, credit, evaluation).
At AgiFid, as at Agitech, the approach stays the same: use AI to save time on repetitive tasks, without ever delegating final responsibility for accounting, tax or HR decisions — that stays a human's job.