The AI training market in Luxembourg is saturated with offers: general catalogues, bootcamps, certification courses, tailored agencies. When you are a training manager comparing several providers, the real question is not the advertised price, but which ones will actually help your teams progress. Here are the 8 objective criteria that separate a good provider from a catalogue that just ticks boxes.
To choose an AI training provider in Luxembourg, compare eight criteria: tailoring to your roles, share of hands-on practice, in-person or remote format, the tools actually taught, follow-up after the session, eligibility for Luxembourg grants (INFPC co-funding, SME Packages AI at 70%), sector expertise and compliance with the EU AI Act.
The landscape of AI training providers in Luxembourg
Before comparing, you need to know who you are dealing with. The public portal lifelong-learning.lu, run by the INFPC, listed in June 2026 nearly 149 artificial intelligence courses offered by around thirty providers. This abundance is misleading: behind the same keyword, “AI training”, you find very different realities.
Roughly speaking, there are four families of players in Luxembourg and the Greater Region:
- Institutions and catalogues such as the House of Training (more than 1,400 courses in its catalogue), Cegos or ABILWAYS, strong on standardised and certified formats.
- Public and support bodies such as the INFPC (portal and co-funding) and Luxinnovation, which guide and support SMEs.
- Schools and bootcamps such as AI2 Education, Level Training, Jedha or KeyJob, geared more towards technical upskilling and qualifications.
- Tailored agencies and firms such as Almera, Retis or AR Digital, which build the training around your real use cases rather than around a fixed programme.
In May 2026, Almera also published a comparison, “TOP 5 AI training providers in Luxembourg”, which puts itself at the top. It is a useful read, but it is a publisher’s ranking, not a neutral grid. The aim of this article is the opposite: to give you the criteria to judge for yourself, whichever provider you consider.
The 8 criteria for choosing your AI training provider
Go through these eight points for each provider on your shortlist. None of them is incidental: it is their combination that predicts whether the training will have a lasting effect or just a pleasant memory.
1. Tailored or catalogue: does the training speak to your roles
Generic AI training shows how to summarise a text or generate an image. Tailored training starts from your real tasks: writing a response to a tender, making an HR report more reliable, sorting client requests. Always ask: will the content be adapted to my sector and my tools, or is it the same material for everyone. This is the criterion that weighs most on return on investment.
2. The share of real practice
Good AI training devotes at least 60% of the time to hands-on practice, on your own cases and your own data. Be wary of days that are 100% demonstration. Ask for the ratio of theory to workshop, the number of participants per session (a small group allows real support) and whether everyone leaves with concrete cases already tested.
3. The tools actually taught, and how they are updated
The ecosystem moves fast. A credible provider teaches the tools your teams will really use: ChatGPT, Claude, Microsoft Copilot, Gemini, and updates its content as new versions come out. Ask the simple question: when was the programme last updated. Material dated 2024 on such a fast-moving subject is a warning sign.
4. The format: in person, remote or hybrid
In Luxembourg, the question of format is also a question of geography. With a large share of Belgian, French and German cross-border commuters, the right provider offers in-person sessions on site, remote sessions and hybrid ones, without lowering the quality of teaching. The mixed format combines the richness of in-room workshops with flexibility for dispersed teams.
5. Follow-up after the training
This is the criterion most often forgotten, and the most telling. Training with no follow-up dissolves within a few weeks. Check what is planned after the session: access to the trainer for several weeks, updated resources, a follow-up meeting, a channel for questions. Without post-training support, new habits do not take root in daily work.
6. Eligibility for Luxembourg grants
Watch out for a common trap: many pages talk about CPF and OPCO, which are French schemes. In Luxembourg, the levers are the co-funding of in-company training run by the INFPC and the SME Packages AI, which fund up to 70% of an AI integration project. A good partner knows these schemes and structures the training so that it is eligible.
7. Sector expertise and local references
Training a team in finance, industry or a services firm does not use the same content. Ask for references from companies in Luxembourg or the Greater Region, examples by profession, and proof that the trainer knows your sector. A use case drawn from your own reality is worth more than ten generic examples.
8. Compliance and data confidentiality
Using generative AI in a company touches on the GDPR and now on the EU AI Act, whose first obligations have applied since 2025. A serious provider also trains your teams in the right reflexes: which data to never paste into a public tool, how to configure privacy settings, which uses are risky. Compliance is part of the skill, not an option.
General catalogue or tailored training: the real trade-off
In short: the catalogue reassures with its price and clarity, while tailored training delivers business impact. For a company project that needs to change practices, tailored training is almost always the best net investment.
Most decisions come down to this trade-off. Here is how the two approaches compare on the points that matter to a training manager.
| Criterion | General catalogue | Tailored training |
|---|---|---|
| Content | Fixed programme, same examples for everyone | Built on your cases and your data |
| Practice | Often theoretical | Workshops on your real tasks |
| Return on investment | Diffuse, hard to measure | Visible on specific tasks |
| Advertised price | Lower per person | Higher, but targeted |
| Follow-up | Rarely included | Support after the session |
As a market benchmark, catalogue AI training often costs between 470 EUR excl. VAT per day (for example at KeyJob) and 1,200 EUR per person for a short session. Tailored training is priced by quote, but should be assessed as a net cost after grants, not as a catalogue price. This is exactly where the funding criterion changes the picture.
i
The reflex to have: never compare two providers on the daily price alone. Compare the net cost after funding, set against the expected impact on your tasks. Tailored training funded at 70% often works out cheaper than a catalogue course paid at full price.
Funding your AI training in Luxembourg, without confusing the schemes
This is where many companies go wrong, by applying French rules to the Luxembourg context. Two schemes really matter in the Grand Duchy:
- Co-funding of in-company training (INFPC): the Luxembourg State co-funds part of the annual training investment of eligible companies. It is cross-cutting support, to be checked with the INFPC according to your situation.
- SME Packages AI: they fund up to 70% of the eligible costs of an AI integration project in an SME, training included, on a project of 3,000 to 25,000 EUR excl. VAT, which means up to 17,500 EUR of support.
Grant to seize
SME Packages AI: your AI training funded at 70%
up to 17,500 EUR
of support for an eligible SME in Luxembourg, which is 70% of an AI integration project of between 3,000 and 25,000 EUR excl. VAT. The training of your teams can be part of the funded project, provided the application is submitted before committing the expense.
→
For details on the amounts, the expenses covered and the procedure, see our page dedicated to the The decision grid to keep to hand
Before signing, run each provider through these questions. If the answer to several of them is vague, keep looking.
- Will the content be built on our roles and real cases, or is it a standard programme.
- What share of the time is devoted to practice on our own data.
- Which tools are taught and when was the programme last updated.
- Which formats are offered (in person, remote, hybrid) for our teams, including cross-border ones.
- What is planned after the training in terms of follow-up and access to the trainer.
- Is the training eligible for INFPC co-funding or the SME Packages AI, and who puts the application together.
- What local references can I be shown in my sector.
- Is compliance (GDPR, EU AI Act) covered in the programme.
Why AR Digital ticks all 8 criteria
AR Digital is a digital agency based in Luxembourg. Our approach is simple: we do not sell a catalogue, we build the training around your real tasks, by function (leadership, marketing, HR, operations), in person or remotely, for your teams in Luxembourg and across the Greater Region.
In practice, we scope the project with you, we train on the tools you will really use (ChatGPT, Claude, Copilot), we plan follow-up after the session, and we structure each programme so that it is eligible for the SME Packages AI or INFPC co-funding. You can explore the content and formats of our
60%+
of the time spent practising on your own cases
70%
funding possible through the SME Packages AI
8/8
selection criteria covered by our approach
Frequently asked questions
How do I choose an AI training provider in Luxembourg?
Compare each provider on eight criteria: tailoring to your roles, share of hands-on practice, the tools actually taught, the formats offered, follow-up after the session, eligibility for Luxembourg grants, sector expertise and compliance with the EU AI Act. It is their combination, not the advertised price, that predicts the real impact of the training.
Which is the best AI training provider in Luxembourg?
There is no best provider in absolute terms: the right choice depends on your sector, your level and your goal. Catalogues such as the House of Training suit standardised courses, while a tailored agency such as AR Digital suits a company that wants to transform its practices on concrete cases.
Catalogue or tailored training: which should I choose for my company?
The catalogue is clearer and cheaper per person, but stays generic. Tailored training costs more on paper, but it is built on your real tasks and produces a measurable return on investment. For a project that really needs to change practices, tailored training funded at 70% often gives the best net cost.
Can AI training be funded in Luxembourg?
Yes, but not through the CPF, which is a French scheme. In Luxembourg, you can use the co-funding of in-company training run by the INFPC, and the SME Packages AI, which fund up to 70% of an AI integration project, training included, up to 17,500 EUR of support.
Should I prefer in-person or remote training?
Both have their place. In-person sessions favour workshops and direct support, while remote sessions reduce logistics costs and suit dispersed or cross-border teams. In Luxembourg, the hybrid format is often the best compromise, provided the quality of teaching is not lowered remotely.
Why is post-training follow-up a decisive criterion?
Because training with no follow-up dissolves within a few weeks. A good provider plans access to the trainer after the session, updated resources and a follow-up meeting, so that new habits take root in daily work. Without this support, the initial investment loses most of its value.









