Artificial intelligence is gradually becoming part of the daily operations of training organizations. But between the European AI Act, legitimate questions around data security, and the ever-present GDPR, it can be hard to find your way. No need to panic: here’s a breakdown of the obligations that apply to you and the best practices to adopt 😉
AI Act: What Does It Change for Training Organizations?
The AI Act is the new European regulatory framework governing the use of artificial intelligence. Its goal: to ensure AI that is safe, transparent, and respectful of fundamental rights.
The regulation classifies AI systems according to their risk level (minimal, limited, high, or unacceptable) and distinguishes two main roles:
- Providers — those who develop or commercialize AI systems.
- Deployers — professionals who use these systems on a daily basis. As a training organization, that’s you.
You therefore have obligations to comply with starting now. But don’t worry: they are manageable, and I’ll walk you through them.
Artificial Intelligence: Your Obligations as a Training Organization
Train Your Teams on AI (mandatory since February 2025)
Article 4 of the AI Act requires you to ensure that your staff have an adequate understanding of AI. They need to understand how the AI tools they use work and what their limitations are, the associated risks (bias, hallucinations), and best practices for use.
Rest assured: no certification is required. A documented awareness program tailored to your tools is sufficient.
Train Your Teams for Free
To meet your AI awareness obligation, the course “Using AI in Your Training Practice” is available for free on the Digiforma Academy and can be completed in under 2 hours. Worth keeping in mind! 😉
Informing Your Contacts (from 2 August 2026)
Article 50 of the AI Act requires transparency around the use of artificial intelligence. Your users must be informed when they are interacting with an AI, and any content generated by artificial intelligence (text, image, audio) must be explicitly identified as such.
In practice, for a training organization: if you use AI to generate a training report or an email for your participants, say so.
Keeping Control Over the Final Decision
AI must never make a decision on its own that has a significant impact on a person. It serves as a preparation and support tool, but the final decision must remain with a human. This is a fundamental principle of the AI Act.
Protecting Everyone’s Personal Data
It’s been a while since you last heard about the GDPR — yet the use of AI remains subject to European data protection regulations. In practice, here are the habits to adopt:
- never enter personal data (from your learners or your clients) into consumer AI tools such as ChatGPT or Copilot;
- check the data processing terms of your professional tools;
- apply the principles of data minimisation and retention limits;
- do not keep AI-generated content longer than necessary.
Maintaining a Critical Mindset
AI-generated output can contain errors. It is essential to systematically review and validate all content before using it. AI is a support tool, not an oracle! 🔮
Non-Compliance with the AI Act: What Are the Penalties?
Failures to meet the obligations set out in the AI Act can result in fines of up to €15 million or 3% of global annual turnover. These amounts are obviously designed with large organizations in mind, but the message from regulators is clear: they will be less lenient than they were in the early days of the GDPR. AI compliance is not optional!
Our Advice: Structure Your Internal AI Usage
Beyond legal obligations, take the time to audit how AI is being used within your organization. To do so, here are some key questions to ask yourself:
- Which AI tools are used in your training organization, even occasionally?
- Who uses them, and for what specific purposes?
- Have your teams received documented awareness training?
- Are your learners informed about the use of AI in their learning journey?
- What data is fed into these tools? Is there a risk of personal data leakage?
- Is there an internal usage policy, even a simple one, to set a framework?
If some answers are still unclear, now is the right time to lay the groundwork!
Did you know that AI is also built into your Digiforma account? Meet Pétronille, our AI Assistant designed 100% for training organizations — join one of our presentation webinars to see her in action!
Meet Pétronille: Digiforma’s AI Assistant Built for Compliance
Pétronille is the AI assistant designed for training organizations, fully integrated into Digiforma. Her mission: to save you time by handling time-consuming tasks on your behalf — such as creating sessions, importing learners, checking your Qualiopi compliance, analyzing your reports… She works like a true teammate, in strict compliance with the law!
AI Hosted in Europe
Pétronille’s technical foundation is built on Mistral technology, a French AI hosted… in Europe! As a result, your data stays on European soil and your data sovereignty is preserved.
Your Data Is Never Used to Train the AI
This is what sets Pétronille apart from consumer AI tools: her model is not trained on your data. Your sessions, learners, and training programs remain strictly confined to your Digiforma account! Our experts improve Pétronille’s capabilities based on their own observations and your feedback — never using your data.
Stay in Control of Every Interaction
In line with the requirements of the AI Act, every interaction with Pétronille is clearly identified and traceable. You always know when you are communicating with the AI, and every action she suggests is summarized before you confirm it.
Pétronille also follows a simple principle: she suggests, you decide. No action is carried out autonomously in your account. She never deletes anything and always asks for your confirmation before acting.
A Strictly Controlled Scope
Pétronille operates exclusively within the boundaries of your Digiforma account and your user permissions. She cannot access data that you yourself do not have access to, and cannot perform tasks on behalf of your contributors that exceed their own permissions.
No Sensitive Decisions Made by the AI
Pétronille does not assess your learners and does not make admission or exclusion decisions. Her role is limited to managing day-to-day tasks: creating sessions, importing learners, Qualiopi checks, report analysis… She handles the tedious work, so you can focus on what matters: training!
Want to Add Pétronille to Your Team?
Good news: she’s included in Digiforma’s 2025-2026 subscriptions! To view your current plan, head to the “My Subscription” page and contact our support team to upgrade it! 👇
