The Diag Data IA is a Bpifrance scheme that co-funds 40% of a data and artificial intelligence diagnosis for French SMEs and mid-caps. Concretely: 8 days of support by an approved expert, billed 10,000 euros excl. VAT, of which 4,000 euros are covered. Your remaining cost is 6,000 euros excl. VAT.
For industrial and services SMEs hesitating to invest in an AI audit, it is currently the cheapest entry point on the market to a professional diagnosis. Here is the complete guide: who is eligible, what the 8 days cover, the typical calendar, and how to apply.
What is the Diag Data IA?
The Diag Data IA is tailored support by a Bpifrance Conseil expert, designed to assess a company's data and AI maturity and identify concrete AI use cases in its operations. It results in a prioritised roadmap covering data, business processes, internal skills and governance (source: francenum.gouv.fr, June 2026).
The scheme is part of the France 2030 plan: 2,000 Diag Data IA engagements are co-funded over 2026-2027. Budgets are therefore limited and served on a first-come, first-served basis.
Who is eligible for the Diag Data IA?
Eligibility conditions, as of 17 June 2026, are the following:
- Be an SME in the European sense, or a mid-cap with fewer than 2,000 employees (the lower threshold is 10 employees).
- Be registered in France with the trade and companies register (RCS) at the date of application.
- Not have already used the same support during the current period.
Micro-companies with fewer than 10 employees are not eligible. For them, a focused AI audit remains possible directly, without subsidy, for a budget comparable to the scheme's remaining cost.
How much does the Diag Data IA cost after subsidy?
The service is billed 10,000 euros excl. VAT. Bpifrance subsidises 40% of it, i.e. 4,000 euros. The remaining cost for an eligible company is therefore 6,000 euros excl. VAT for 8 expert days. For comparison, a day of AI consulting commonly bills 1,200 to 1,800 euros on the French market: the scheme brings the expert day down to 750 euros.
What do the 8 days of support cover?
The expert's 8 days are spread over 3 months maximum and cover four blocks:
- Data and AI baseline: data maturity, tools in place, existing usage.
- Use-case identification: business workshops to spot the value deposits.
- Prioritisation: quantifying and ranking use cases by expected return.
- Roadmap: sequenced rollout plan, with associated skills and governance.
The content largely overlaps with a classic AI audit method. To understand each step in detail (and the traps to avoid), read our guide AI audit: the method to prioritise the right use cases.
The typical calendar of a Diag Data IA
- Week 0: application, eligibility validation by Bpifrance and framing with the assigned expert.
- Weeks 1-2: baseline, team interviews, collection of real volumes (2 to 3 expert days).
- Weeks 3-6: use-case workshops per department, gain quantification, feasibility checks (3 to 4 days).
- Weeks 7-10: handover of the prioritised roadmap, implementation funding plan (1 to 2 days).
How to apply?
- Check your eligibility (size, RCS registration) and prepare your company information.
- Submit your application on Bpifrance's online platform, under Diag Data IA.
- Bpifrance validates eligibility and assigns an approved expert to frame the engagement.
- The 8-day support runs over 3 months maximum and ends with the roadmap handover.
After the diagnosis: funding implementation with IA Booster
The Diag Data IA funds the diagnosis, not the build. For the next step, Bpifrance operates IA Booster France 2030: support for the concrete implementation of the identified use cases (data, development, integration). The two schemes are designed to chain: the Diag roadmap serves as the entry file for IA Booster.
That is the logic we recommend in every case, subsidy or not: a diagnosis only has value if the identified use cases actually get built. Budget the first quick win's implementation at the same time as the diagnosis.
Diag Data IA or tailored AI audit: how to choose
The Diag Data IA is a great first structuring move: framed format, unbeatable price thanks to the subsidy, and the internal credibility of the Bpifrance label. A tailored AI audit keeps the edge when the need is sharper: a scope focused on one department, a demanding sector (private equity, healthcare), or the intent to move straight into deploying the quick wins with the same team.
In both cases, the provider-selection criterion stays the same: demand references of real deployments, not only consulting references, and talk to a client supported more than 6 months ago.
