EDIHs and Swiss AI Centre: How Swiss SMEs Access Free AI Programmes

Switzerland has had a national AI action plan since July 2026 - but the real question for your business is no longer whether to engage with it, but how to use the infrastructure it already funds. European Digital Innovation Hubs (EDIHs), the Swiss AI Centre, the Canton Zurich AI Innovation Programme and Innosuisse funding schemes already offer Swiss SMEs free consulting, test infrastructure, access to AI engineers and up to 50 percent project funding. Access is open - it's simply underused, because most SMEs don't know where to start.

The Problem These Programmes Solve

A study by HWZ and Swisscom covering 123 Swiss SMEs (September/October 2024) identified three main barriers to AI adoption: a shortage of skilled talent, regulatory uncertainty, and a lack of internal expertise. EDIHs address exactly these three points structurally - not as a subsidy in the classic sense, but as external capacity you tap into temporarily before committing your own investment.

Four Access Points at a Glance

• EDIHs and the Swiss AI Centre: consulting, test infrastructure and regulatory guidance, coordinated at cantonal and national level
• Canton Zurich AI Innovation Programme (SDSC): a one-year, free programme for SMEs culminating in concrete pilots
• Innosuisse funding: up to 50 percent of project costs for AI initiatives run with a research partner
• AI Days 2026 (HES-SO): a three-day practical workshop series on ML fundamentals, generative AI and computer vision

European Digital Innovation Hubs and the Swiss AI Centre

Switzerland's national AI roadmap explicitly names EDIHs as SME support infrastructure. In practice, EDIHs provide four services: consulting on identifying suitable use cases and building a roadmap, access to test infrastructure such as sandbox environments, compute capacity and trial licenses, access to specialists including AI engineers, data scientists and domain experts, and regulatory guidance on the EU AI Act, Switzerland's revised data protection act (revDSG), and sector-specific requirements. For an SME without its own data science team, this is often the only realistic way to evaluate a first pilot properly.

Canton Zurich AI Innovation Programme

Launched by the Canton of Zurich's Business and Economic Development office together with the Swiss Data Science Center (SDSC), the programme runs for one year, is free for participating SMEs, and follows three stages: demystifying AI, developing prototypes, and advancing promising pilot projects. The first cohort launched in September 2025, selecting 12 SMEs from 28 applicants - a ratio that shows preparation pays off. The next cohort starts in September 2026, with an information session scheduled for 30 June 2026, 13:00-13:45 CEST. Applications run through an online form.

Innosuisse: Up to 50 Percent Project Funding

Innosuisse funds up to 50 percent of project costs for SMEs and startups developing AI products or services in collaboration with a Swiss research institution. Flagship calls exist for research-intensive AI projects in life sciences, typically funding CHF 100,000 to 1 million. The Swiss Accelerator targets SMEs with above-average innovation potential through a competitive process. At cantonal level, Vaud complements this with DigiBoost vouchers of CHF 10,000 for SMEs with fewer than 50 employees, financed through the New Regional Policy fund.

AI Days 2026 by HES-SO

From 23 to 25 March 2026, AI Days 2026 takes place in Martigny and Fribourg, organised by HES-SO and the Swiss AI Centre. With more than 350 expected participants, the three days cover workshops on ML fundamentals for SMEs, generative AI in business processes, computer vision for industrial applications, methods from concept to prototype, and an overview of funding options. Experts involved include Francesco Carrino (HEI Valais, Local Chair), Thilo Stadelmann (ZHAW), and Sébastien Rumley and Jean Hennebert (both HEIA-FR). Partners include the Eksperience Festival, Alp ICT, the economic promotion office of Canton Fribourg, Pure Storage, Exoscale, and Valais Economic Promotion.

How to Apply - Step by Step

1. Identify a specific internal use case - not 'AI in general' - a process with measurable effort that could be automated or supported
2. Check which programme fits: EDIH consulting for initial orientation, the Canton Zurich AI Innovation Programme for a structured year-long engagement, Innosuisse for a project with a research partner
3. Submit your application through the relevant online form on time - for the Canton Zurich programme, before the September cohort start; for Innosuisse, depending on the specific call
4. Prepare for a selection conversation: evaluators typically assess clarity of the use case, readiness to implement, and availability of internal resources for the collaboration
5. Plan for internal involvement - even a free programme requires time from your team, particularly for data access and domain knowledge

What You Actually Get

• Free consulting on identifying and prioritising use cases
• Access to test infrastructure: sandbox environments, compute capacity, trial licenses
• Direct access to AI engineers, data scientists and domain experts without hiring them
• Regulatory guidance on the EU AI Act, revDSG and sector-specific requirements
• Up to 50 percent funding of project costs through Innosuisse when working with a research partner

Early Results from the First Cohort

The first cohort of the Canton Zurich AI Innovation Programme produced two collaborative open-source projects: a RAG pipeline and a time-series forecasting model, both released under an MIT license on GitHub. This matters because it shows that a free, publicly backed programme can generate not just individual pilots, but reusable building blocks for the wider SME landscape.

The inaugural cohort represents a robust group poised to drive meaningful and lasting AI innovation.

Why Now Is the Right Moment

Falling model costs are lowering the technical barrier to entry, while EDIHs and the programmes described here simultaneously lower the knowledge and capacity barrier. Together, these two effects noticeably shift the timeline for SMEs: what was only realistic with an in-house data science team two years ago is now accessible through time-limited external support.

It's worth noting that a free programme doesn't replace a permanent structure. Once a pilot succeeds, you need an internal or externally coordinated function that takes over prioritisation, governance and scaling - the programmes are the entry point, not the answer for the next five years.