How to choose AI training for your company or leadership team: 7 questions to ask before you sign

By Patrik Rouault · Published October 27, 2026 · FUNDAE information checked October 27, 2026

Before you hire a provider to train your team in AI, ask them seven concrete questions: what level they start from, what cases you will work on, what stays configured at the end and who delivers the training, among others. The answers tell you whether you are buying training that carries into daily work or a generic course that never gets used.

At Coach IA Pro, we train teams and executives in AI. These are the questions we would ask any provider, including us.

Who this guide is for. Leaders, HR managers and partners of international companies and firms based in Spain or elsewhere in Europe who are about to choose AI training for their team or leadership team.

1. Before comparing providers: what you want to achieve

Before asking for proposals, clarify three points. That way you will compare answers to the same question.

  • Who to train: the leadership team, one department or the whole staff. The goals and the pace are not the same.
  • For which tasks: drafting, analyzing documents, preparing meetings, answering clients. The more specific they are, the easier it is to judge whether the program addresses them.
  • With which tools: the ones you use every day (Google Workspace or Microsoft 365) and the AI accounts you may already have (ChatGPT, Claude or Gemini) without feeling you get the most out of them; and also the ones you are willing to pay for.

If you are an executive and want to train yourself first, the same questions apply to one-to-one training.

2. The 7 questions to ask before you sign

QuestionWhy it mattersWhat answer to expect
1. Do they start from each person's actual level and group people by level, or give everyone the same course?In the same team, some people have never used AI and others use it every day. A single course bores some and loses others.A prior assessment (questionnaire or interviews) and an explanation of how groups are formed.
2. Will you work on your own cases and documents, or on generic examples?What is practiced on generic examples rarely makes it into daily work.A request for your team's tasks or documents before starting, with the necessary confidentiality precautions.
3. What stays configured at the end: projects, assistants, usage rules?Without configuration, everyone starts from scratch again.A concrete list: configured projects or assistants, saved instructions, usage rules.
4. Who delivers the training, and do they use these tools every day in their own work?The tools change every few months: someone who uses them daily knows what works today.The name of the trainer and examples of their own use.
5. Are they independent of AI vendors? How is the tool chosen?A provider tied to a vendor tends to recommend that vendor's tool.A clear statement of any commercial ties, and a tool chosen for your tasks and your environment.
6. How is confidentiality protected?During training, you work with company information.A non-disclosure agreement if you ask for one, business plans for working with real data, and a data processing agreement (GDPR Article 28) if the provider accesses personal data.
7. Is there follow-up after the training? How do you measure whether the team uses it?The difficulties appear back in daily work, not during the session.Planned follow-up sessions and a simple criterion: which use cases are still active a few weeks later.

1. Do they start from each person's actual level and group people by level, or give everyone the same course?

Ask how each person's level is known before the first session. If the answer is "we'll see on the day," the program will be the same for everyone. A group at a similar level moves faster and is more willing to ask questions.

2. Will you work on your own cases and documents, or on generic examples?

Good training starts from your tasks: the monthly report, sales proposals, client replies. Ask for an example exercise adapted to your sector before you sign.

3. What stays configured at the end: projects, assistants, usage rules?

Learning to write good instructions is useful, but what lasts is what stays configured: a project with your reference documents, an assistant for a recurring task, shared usage rules. Ask for the proposal to list them.

4. Who delivers the training, and do they use these tools every day in their own work?

Ask who will be in the room, not just who signs the proposal. Ask them to show how they use AI in their own work: if they can't, that is a signal.

5. Are they independent of AI vendors? How is the tool chosen?

Claude, ChatGPT and Gemini are not strong at the same tasks, and your company's environment matters. An independent provider should be able to explain why they recommend a tool for your case. We compare the three in Claude vs ChatGPT vs Gemini for executives and companies.

6. How is confidentiality protected?

Ask which accounts will be used during the training and what happens to the documents you share. Free or personal versions of the tools do not offer the same terms as business plans. We explain this in Is it safe to use AI with company data?.

7. Is there follow-up after the training? How do you measure whether the team uses it?

Ask what happens after the last session. A simple criterion: check, a few weeks later, which use cases are still active, which have been dropped, and why.

3. How we answer these questions

  1. We start from each person's actual level, with a questionnaire or individual interviews. We form groups of similar level when the size and organization of the group allow it, including for the one-day workshop; it is not always possible.
  2. We work on your tasks and documents, with the precautions in question 6.
  3. In the tailored AI program, what stays is the use-case mapping, the implementation plan and whatever was configured during the use-case workshops.
  4. The training is delivered by Patrik Rouault, founder of Coach IA Pro, who uses Claude, ChatGPT and Gemini every day in his work.
  5. We have no commercial agreement, commission or affiliation with any AI vendor.
  6. We sign a non-disclosure agreement when the client asks for one.
  7. After each training, we offer follow-up sessions to answer questions on real cases and adjust the configuration: the training doesn't end on the last day.

The formats are described on AI training and consulting for companies and leadership teams.

4. Red flags

  • Time savings promised in figures before anyone has seen your work.
  • The same program for everyone, with no prior assessment.
  • No deliverable at the end, just a slide deck.
  • Free or personal accounts used with client data.
  • A trainer who can't show examples of their own use.
  • A single tool presented as the answer to everything.

5. Checklist before you sign

Copy this list to compare two or three proposals:

  • ☐ Each person's level is assessed before starting.
  • ☐ Groups are organized by level, or the proposal explains why not.
  • ☐ We work on our own tasks and documents.
  • ☐ The proposal lists what stays configured at the end.
  • ☐ We know who delivers the training and how they use AI in their work.
  • ☐ The provider states whether they have ties to any AI vendor.
  • ☐ Confidentiality is covered in writing.
  • ☐ Follow-up is planned after the training.
  • ☐ The proposal does not promise savings in figures before seeing our work.

6. If you run a professional services firm

Law, accounting and advisory firms work with client documents covered by professional secrecy or confidentiality commitments. On top of the seven questions, add two:

  • Who can see the documents used during the training, and which tool and plan are used?
  • Does the training include an internal usage protocol: what can be shared with AI and who reviews the results?

Each profession's ethical rules may add obligations of their own: the provider should know them, and the training should be adapted to them. We explain how we work with firms in AI for professional services firms.

7. AI training and the EU AI Act

Since February 2, 2025, Article 4 of the EU AI Act has required organizations that use AI to take measures to support their staff's AI literacy. Since July 2026, it no longer mandates a specific level, but the obligation remains.

Further reading: AI acceptable use policy: a template for small businesses.

8. When we are not the right fit

  • If your priority is training fully covered by a public funding scheme, you are better off with a provider certified for those schemes.
  • If your company has chosen Microsoft Copilot: we don't train on Copilot; we work with Claude, ChatGPT and Gemini, the tools we use every day.
  • If you need to train hundreds of people at once, a training company with more trainers will be a better fit.
  • If you are looking for a self-paced online course with no support, a course platform will suit you better: our training is live and built on your work.

9. Frequently asked questions

How long does AI training for a team take?

It depends on the format. A one-day workshop gets you started on concrete cases; a program over several weeks lets you configure the tools and support people as they put it into practice. Rather than the number of hours, check that it includes practice on your cases and follow-up.

Can a company based in Spain fund the training through FUNDAE?

In Spain, companies that pay vocational training contributions have an annual training credit from FUNDAE, the state foundation for training in employment. The credit is applied as a reduction in social security contributions once the training has taken place. It depends on the vocational training contributions paid the previous year and on headcount. Companies with 1 to 5 employees have a guaranteed minimum credit of €420; from 6 employees, the company must co-fund part of the cost. FUNDAE can cover part of the cost, but it shouldn't decide your choice: choose the training that answers your questions first, then check eligibility with your gestoría (payroll and tax adviser) before you sign. FUNDAE offers a credit simulator (in Spanish). In other countries, ask about the national scheme.

Is in-person or remote training better?

Both work. Team workshops often benefit from being in person, because people practice together on the same cases; individual interviews and follow-up work well by video call. More than the format, check that the training is built on your tasks.

Should we start with the leadership team or the whole team?

Often with leadership: they decide which tools are bought and which rules apply, and their example matters. Then the team, by level or by function. For a first individual step, see our practical AI guide for executives.

Which AI tool should we choose before training?

You don't need to decide beforehand: a good provider helps you choose based on your tasks and your working environment, for example Google Workspace or Microsoft 365. We compare the three main tools in Claude vs ChatGPT vs Gemini for executives and companies.

Want to compare with a concrete proposal? Book a free 30-minute intro call, in Madrid or remotely, in English, Spanish or French. After the call, we send you a tailored proposal. You can also email us at patrik@coachiapro.com. Prefer to start with your own level? Take the AI skills test.

Sources (accessed October 27, 2026)