Switzerland is not an EU member and does not adopt the AI Act. For many Swiss companies the Regulation is binding nonetheless, because it attaches to the market and to effect, not to the registered office. At the same time, Switzerland is developing its own approach. Those active in both areas need to know both.
When the AI Act covers Swiss companies
The AI Act applies to providers placing AI systems on the market or putting them into service in the EU, regardless of where they are established. It also applies to providers and deployers in third countries where the system’s output is used in the EU. Typical cases: a Swiss SaaS provider sells a product with an AI feature to customers in Germany — it is a provider under the AI Act. A Swiss company uses an AI system to assess applications for a position at an EU subsidiary — it may be covered as a deployer. An internal tool used purely in Switzerland with no EU link, by contrast, is not covered.
What it means in practice
Swiss companies in scope face the same obligations as EU companies: AI literacy under Art. 4, transparency obligations under Art. 50 and, for high-risk systems, the full set of obligations from December 2027 or August 2028 respectively. Providers of high-risk systems and of general-purpose AI models established outside the EU must additionally appoint an authorised representative in the EU to act as the contact point for authorities. Which role a company has per system is described in the glossary entry on provider and deployer.
The Swiss path
In February 2025 the Federal Council decided to ratify the Council of Europe’s AI Convention and to make the necessary amendments to Swiss law, as far as possible sector by sector rather than through a comprehensive AI act on the EU model. Where changes are needed they are to be made in existing laws, complemented by non-binding measures. A consultation draft is planned for the end of 2026. For companies this means: Switzerland will not create a duplicate of the AI Act, but its own regulation that will differ on individual points.
What already applies: the FADP
Independently of any AI regulation, the revised Federal Act on Data Protection applies to AI applications as soon as personal data is processed. The Federal Data Protection and Information Commissioner has stated this explicitly. Concrete duties follow: transparency about the purpose and type of processing, information on automated individual decisions and the data subject’s right to state their position, a data protection impact assessment where risk is high, and the rules on disclosure abroad when an AI provider processes data outside Switzerland.
Handling dual requirements in practice
For companies with EU business it is usually efficient to align with the stricter standard and add Swiss specifics. A joint AI inventory with role, risk class under the AI Act and data protection classification under the FADP and GDPR covers both worlds. AI literacy measures under Art. 4 simultaneously serve the responsible handling of personal data under the FADP. And transparency notices for chatbots are good practice in Switzerland too.
What to do now
1. Check which AI systems have an EU link, as a product or through use of their output.
2. For these systems, determine role and risk class under the AI Act, for example with our high-risk classification checklist.
3. For high-risk systems and GPAI models, clarify the question of an EU authorised representative.
4. For all AI applications processing personal data, implement the FADP requirements.
5. Follow the Swiss consultation and plan for adjustments.
The bigger picture is in our AI Act 2026 guide. We support implementation through our AI Act consulting, based in Zug with an eye on both legal areas.