Switzerland has unveiled an ambitious roadmap to strengthen its position as a global leader in Artificial Intelligence. Through a comprehensive set of actions spanning education, research, infrastructure, data governance, and regulation, this Switzerland AI Roadmap aims to transform Switzerland’s existing strengths into a coordinated national strategy capable of driving innovation, competitiveness, and societal benefit in the age of AI.
The AI Action Plan for Switzerland, coordinated by digitalswitzerland, provides on the one side a pragmatic entry point to the Swiss endeavours on AI, and on the other side a map of the overall AI ecosystem in the Federation, aiming to strategically break silos and harmonise practices at the federal level and across the cantons.
Building on its long tradition of neutrality and its central role in international cooperation and scientific research, hosting organisations such as the United Nations and CERN, Switzerland is combining existing public and private initiatives with a new strategic framework and targeted actions to secure a leading position in the global development of AI technologies.
Scaled AI education and literacy
The plan is structured to address five key areas, namely Scaled AI Education and Literacy, World Class Research and Innovation, Resilient Digital Infrastructure, AI-Ready Data and Smart AI Governance. This key area in the Switzerland AI Roadmap focuses on the empowerment of the Swiss population to acquire a deeper understanding of AI technologies, enabling it to make competent and ethical use of AI, as well as to join the public debate and make informed decisions. The actions in this area will focus on providing tailored training both in the educational system and for enterprises, with the objective of supporting the workforce adaptation to AI. The actions include the launch of Switzerland’s public campaign for the responsible use of AI, planned for 2027 and targeting the general public.
Two other measures underlying this area of intervention revolve around the Swiss educational system. The first, “Strengthen Swiss-wide AI Education and Literacy”, will target secondary- and university-level trainers, as well as policymakers in the education field, aims at coordinating AI education and at supporting best-practice exchange. The following action, “AI Skilling for Teachers”, specifically aims at providing secondary-level and university-level teachers with the tools to both increase AI literacy among students and to use AI to teach.
The last two actions identified in this key area look into means to increase the adoption of Artificial Intelligence solutions in SMEs, specifically through a “Playbook for AI Adoption” and to establish a “Certificate for Trustworthy AI Use” in the form of micro-credentials as a means to start the up-skilling of the workforce in Switzerland, with a shared framework that can be used across all enterprises. This is planned to be rolled out by digitalswitzerland.
World-class research and innovation
This key area focuses on ensuring Switzerland stays at the forefront of research and innovation, with specific reference to AI. This key area includes action to foster integration of AI technologies in the Swiss industry, fast-tracking AI to practical applications, and contributing to the Swiss economy.
The key actions in this domain that are in the Switzerland AI Roadmap include the establishment of a Swiss Centre for Artificial Intelligence Research and Innovation, nick-named “CERN for AI”, to Deepen International AI Research Collaboration, and to develop Swiss EDIHs (European Digital Innovation Hubs) for AI SMEs support.
From a policy perspective, the plan also aims to deploy “National AI Innovation and Regulatory test environments” (Sandboxes), to allow both small and large actors in the AI ecosystem to experiment in an ecosystem protected from legal liability, as well as informing the development of tailored AI policies which ensures at the same time an adequate level of safeguards, while not regulating AI innovation to a halt.
Finally, this key area includes specific focus on “Swiss AI models”. This action point aims at encouraging the adoption of Swiss AI models, including for instance Apertus (https://apertvs.ai/), thus ensuring the use of sovereign, trusted technology, especially in critical sectors.
Resilient digital infrastructure
This key area aims at ensuring the Swiss digital infrastructure is up-to-the-challenge in a context of growing demand for computation, driven by the increasing adoption of AI solutions. This key area is supported by five underlying actions.
Procurement standards for digital infrastructure
The definition of procurement standards to manage public digital infrastructure dependencies aims to ensure that harmonised and actionable guidelines are available to public offices when identifying and choosing providers of digital services.
While this action in the Switzerland AI Roadmap is partly covered by the existing and mandatory “Schutzbedarfsanalyse” [Protection needs analysis], defined in Art.16, Bundesgesetz über die Informationssicherheit [Federal Law on Information Security], and by the National Strategy for Critical Infrastructure Protection, it will be further expanded by the provision of harmonised operational metrics and enforcement aids.
AI energy efficiency and transparency
The second action aims to increase AI energy efficiency and transparency through the definition of a labelling and disclosure framework for AI infrastructure.
This will build upon the SDEA (Swiss Datacentre Efficiency Association) label.
Forecasting compute and electricity demand
Extending the focus on energy, the Resilient Digital Infrastructure area also encompasses an action aimed at forecasting AI compute and electricity demand in Switzerland, with a view to forecast and plan to satisfy both the compute requirements expected in the coming years, and the subsequent energy demand.
Framework conditions for data centre development
This is further supported by the action aimed at defining the ideal framework conditions for data centre development, which is expected to establish a regulatory and spatial planning framework to support efficient and sustainable data centre development.
e-ID and trust infrastructure
Lastly, the area identifies the need to support the introduction of the e-ID and further development of the trust infrastructure. This action aims at ensuring that identification of Swiss citizens can happen safely, with specific reference to those living abroad.
AI-ready data
The first action listed in the plan is the aim to define and provide an assessment framework for AI-ready data. This responds to public and private actors seeking to share their data for AI use, and for actors who seek to consume the data for AI. This aims at defining a shared understanding around AI-ready data, the assessment criteria to evaluate AI readiness and the development of guidelines for the provision and consumption of AI-ready data.
The second action identified, “Promote secondary use of data to train or run AI system(s)”, is closely related to the latter. This is expected to be implemented through the definition of a legal framework for enabling secondary data use, specifically for AI, the definition of metadata and interoperability frameworks, as well as support for data sharing ecosystems promoting secondary use of data in critical domains.
The third action in the Switzerland AI Roadmap refers to interoperability and aims to promote standards and accessible platforms for public sector data across federal, cantonal, and municipal levels. This primarily aims at improving interoperability via the definition of metadata standards, starting with “minimal viable metadata” which can be adopted in each sector or industry, and the definition of standards to harmonise data as well as the extension of data platforms to support AI systems.
The AI-Ready Data area also aims to complete the establishment of the international Trusted Data Observatory in Geneva. The process has already started in 2025 with stakeholder identification and is planned to be completed in 2030, with a broader onboarding of actors in the sector.
Smart AI governance
The Smart AI Governance key area identifies in the Switzerland AI Roadmap areas of intervention at a regulatory and policy level. The area defines three key actions, starting from the development of a toolkit for responsible AI self-commitments, a set of non-binding guidelines to inform the development of sector-specific codes of practice.
Secondly, National AI Innovation and Regulatory test environments, implemented as a combination of regulatory and technological sandboxes to allow for experimentation and innovation in a safe and monitored environment. This approach also supports the development of adequate policies when exploring novel and not fully explored issues.
Finally, yet importantly, the Smart AI Governance area plans to introduce a Digitalization Article into the constitution, which would allow the definition of provisions on essential technological elements at federal level, reducing the risk of fragmentation across the cantons.
What is in it for businesses?
For Swiss businesses of all sizes, the AI Action Plan’s value proposition is primarily about enabling conditions rather than direct cash handouts — though there are concrete funding angles worth knowing. The headline figure comes from the Google/Implement Consulting Group economic study commissioned alongside the Plan: AI adoption is projected to contribute around CHF 15 billion annually to R&D-driven innovation in Switzerland by 2034, the logic being that companies which move early stand to capture a disproportionate share of that upside.
For any business, the Plan’s most immediately useful pillars are data access and governance.
By aligning business, research, and the public sector, the Plan aims to make high-quality data a reliable and accessible foundation for the next generation of AI-based business models. On governance, Federal Councillor Rösti has been explicit: “We are focusing on smart, agile rules instead of bans”.
Switzerland currently has no dedicated AI legislation, relying on existing legal frameworks and sector-specific regulations, which, for companies building or deploying AI products, means lower regulatory overhead than across the border.
Despite measurable benefits, 52% of Swiss companies expect AI returns within one year, above the European average, and 73% of Swiss executives expect significant revenue contribution by 2030.
Scaling remains limited by challenges such as inadequate system integration, inability to identify viable use cases, lack of AI literacy, and data protection and security concerns. The Plan directly targets these blockers.
Opportunities for SMEs and startups in the Switzerland AI Roadmap
For SMEs and startups specifically, the picture is more concrete.
On compute, the Swiss AI Initiative’s Alps supercomputer, the infrastructure backbone cited throughout the Plan, is accessible via regular open compute calls, and the initiative provides transparent and open software, models, and data releases, enabling their trustworthy use by various Swiss stakeholders, including SMEs and start-ups.
The open-source Apertus LLM (trained on Alps by ETH Zurich, EPFL and CSCS) gives smaller businesses a sovereignty-respecting, compliance-friendly foundation model to build on without paying Big Tech API costs or surrendering data to foreign cloud infrastructure.
On funding, the Switzerland AI Roadmap does not create new grant instruments of its own, but it explicitly works in concert with existing ones.
Innosuisse covers up to 50% of project costs for SMEs and startups developing AI-powered products or services in collaboration with Swiss research institutions.
The Swiss Accelerator is a competitive funding programme aimed at individual start-ups and SMEs carrying out innovation projects with above-average innovation potential.
Cantonally, programmes like Vaud’s DigiBoost offer digitalization vouchers up to CHF 10,000 for SMEs with fewer than 50 employees, and the New Regional Policy Fund provides canton-level support for digital transformation.
For R&D-heavy AI projects in life sciences, Innosuisse Flagship calls offer up to 50% of project costs (typically CHF 100,000–1,000,000), provided the SME or startup partners with a research institution.
The Plan also explicitly promises digitalswitzerland training programs for SMEs, often co-funded, addressing the literacy gap that currently blocks adoption.
One caveat: the Switzerland AI Roadmap does not yet specify binding commitments or enforcement mechanisms, and most of its promised actions remain “rolling” and voluntary.
Businesses should therefore treat it as a signal of direction and a coordination platform, not a guarantee of specific support.
The most reliable near-term channel for funded AI projects remains Innosuisse, with the Plan adding political momentum and the prospect of expanded open data and compute access as it matures toward the 2027 Geneva summit.
