1. What is the understanding or definition of AI in your jurisdiction?
In Switzerland, there is no formal legal definition of 'artificial intelligence' codified in a specific AI law. Instead, the Swiss Federal Government adopts an approach which is both functional and technology-neutral in its treatment of AI. The government's focus is directed towards the operational mechanisms of AI systems and the risks they pose, as opposed to adhering to a rigid technical definition.
In a report published in February 2025,595 the Federal Council deliberately opted against adopting a codified legal definition, emphasising that AI should be regulated through existing sectoral legislation (eg, data protection, financial supervision and product safety) as opposed to terminology specific to the technology.
That said, the Competence Network for Artificial Intelligence (CNAI), part of the federal administration, has issued an AI terminology document to promote uniform language.596 In it, artificial intelligence is described as 'building or programming a computer to do things that normally require human or biological skills (intelligence)'.
2. In your jurisdiction, besides legal tech tools (ie, law firm or claim management, data platforms, etc), are there already actual AI tools or use cases in practice for legal services?
As of 2025, Switzerland's legal sector has firmly moved into practical AI deployment. Approximately 50 per cent of financial institutions actively use AI according to the Swiss Financial Market Supervisory Authority's (FINMA) 2025 survey,597 and large Swiss-based as well as international firms deploy AI for contract analysis, due diligence and legal research. The market is characterised by both international platforms adapted to Swiss requirements – such as multilingual support and data residency – and locally developed solutions specifically designed for Switzerland's legal landscape.
Strict data residency in Switzerland is a recurring requirement in the Swiss legal market (see the response to Question 9, below). It is increasingly met not only by local hosting providers but also by Swiss-based data centres of large international hyperscalers, enabling more international, legal specific AI tools to offer Swiss data residency aligned with client and regulatory expectations.
Several Switzerland based startups are also developing AI enhanced legal research tools tailored to Swiss law. However, these offerings remain at an early stage, with limited market traction compared to more established platforms.
As for specific use cases, contract intelligence platforms see the most widespread adoption, with Swiss-developed tools being used for contract analysis and regulatory compliance. AI solutions integrated into knowledge management enable users to search internal knowledge bases while conducting AI powered research and document review, including in German- and French language environments. There is also an increase in the use of AI-powered solutions in the eDiscovery and investigations realm.
Swiss legal publishers, most of whose content remains behind paywalls, are also beginning to combine authoritative national legal materials with AI-based functionalities.
3. If yes, are these AI tools different regarding
-
independent law firms;
-
international law firms; and
-
in-house counsel; and what are these differences?
In practice, international law firms with Swiss offices and larger in-house legal teams tend to be earlier adopters of AI, mainly because they can leverage existing group-wide IT infrastructure, budgets and governance – especially as they gained know-how from jurisdictions that are ahead of Switzerland in general AI adoption.
Independent Swiss firms – especially small and mid-sized ones – more often use AI in a pragmatic and less standardised way, typically as a drafting and summarising aid on the level of individual lawyers or practice groups. In-house teams, by contrast, are a key target group for AI tools that support fast contract analysis and policy-based risk triage, with the objective of handling more routine contract work internally and reducing reliance on external counsel for high-volume but legally standardised matters.
It is noteworthy that some firms are compelled to maintain dual infrastructures – on‑premise solutions without AI and cloud-based solutions with AI functionalities
- in order to reconcile differing risk profiles and regulatory expectations across their client base.
4. What is the current or planned regulatory approach on AI in general?
Regulation of AI in general is still in development at the time of this writing.
In its report published in February 2025,598 the Federal Council decided to steer AI regulation toward three main goals: strengthening Switzerland's innovation capacity; protecting fundamental rights (including economic freedom); and building public trust. Rather than creating a broad, cross-sector 'AI Act' like the European Union, Switzerland will adopt sector-specific regulation where needed and reserve cross-sector regulation for legal areas where fundamental rights, in particular, are at play (eg, data protection).
Simultaneously, Switzerland is poised to ratify the Council of Europe's Convention on Artificial Intelligence. The Federal Department of Justice (DFJP), in collaboration with other agencies, has been assigned the responsibility of drafting a legislative proposal that is scheduled to be submitted for consultation by the conclusion of 2026. The proposed bill will entail the implementation of requisite legislative amendments to ensure adherence to the principles laid out in the EU's AI Convention. It is anticipated that this will encompass the introduction of targeted provisions pertaining to pivotal issues within the realm of AI, including transparency, data protection, non-discrimination and the establishment of oversight mechanisms.
In parallel to formal regulation, the government continues to use soft-law instruments. A notable example is the AI Guidelines issued by the Federal Council on 25 November 2020, which set out seven key principles for AI use in Switzerland.599 These include putting people first, ensuring transparency/explainability, accountability for AI outcomes, safety and security, etc. The Federal Office of Communications (DETEC/OFCOM) will also develop a non-binding governance plan by end the of 2026, including voluntary codes or self-declaration mechanisms.600
While federal regulation is still in development, sub-federal authorities are adopting binding directives that set minimum standards for how public employees may use online AI tools in their daily work. These rules primarily define which use cases are allowed, which data may not be entered into external AI systems and how AI-generated outputs must be checked and used under human control.
To view the full pdf, click here.
Footnotes
595. DFJP, Basic legal analysis within the framework of the current state of regulatory approaches to artificial intelligence, https://www.bakom.admin.ch/dam/fr/sd-web/4HwAbRI-HeC1/analyse_juritisch.pdf (French version); https://www.bakom.admin.ch/dam/de/sd-web/4HwAbRI-HeC1/analyse_juritisch.pdf (German version), accessed 13 January 2026.
596. Federal Department of Home Affairs, Terminology: Competence Network for Artificial Intelligence (CNAI), December 2023, https://cnai.swiss/wp-content/uploads/2024/08/CNAI_Terminologie_E_2_1.pdf accessed 13 January 2026.
597. ‘FINMA survey: artificial intelligence gaining traction at Swiss financial institutions', FINMA, 24 April 2025 https:// www.finma.ch/en/news/2025/04/20250424-mm-umfrage-ki accessed 13 January 2026.
598. See n1, above.
599. Federal Council, ‘Guidelines on Artificial Intelligence for the Confederation' https://www.sbfi.admin.ch/dam/en/ sd-web/-uy97trD1VMc/Leitlinien%2520K%C3%BCnstliche%2520Intelligenz%2520-%2520DE%5B1%5D.pdf accessed 14 January 2026.
600. Digital Switzerland Strategy, ‘Action plan' https://digital.swiss/en/action-plan accessed 14 January 2026.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
