Switzerland is one of the most transparent countries on earth when it comes to government data. Federal laws, cantonal legislation, court decisions, regulatory publications, commercial registry entries, official gazette notices: all published through official government channels. Much of it structured. Much of it machine-readable. Almost all of it free.
And almost nobody is doing anything useful with it.
What Switzerland Actually Publishes
The scope of Switzerland’s open government data is remarkable. Most professionals interact with small slices of it through specialized portals, but few grasp the full picture.
Federal legislation. Fedlex (fedlex.data.admin.ch) publishes every federal law, ordinance, and international treaty in all three official languages plus English. The data is available as structured XML with detailed metadata: SR numbers, enactment dates, amendment histories, and relationships between provisions. There are over 4,900 federal enactments, comprising tens of thousands of individual articles.
Cantonal legislation. Each of Switzerland’s 26 cantons publishes its own legislation through cantonal law databases. The formats vary (some structured, some not), but the data is publicly available. Across all cantons, this represents approximately 23,000 additional legal enactments with hundreds of thousands of individual provisions.
Court decisions. The Federal Supreme Court (BGer) publishes all decisions through its official portal. The Federal Administrative Court (BVGer) and the Federal Criminal Court (BStGer) do the same. Many cantonal courts publish decisions as well. In total, Swiss courts have published over a million searchable decisions spanning decades of jurisprudence.
SHAB (Swiss Official Gazette of Commerce). Every company formation, dissolution, merger, board change, capital modification, and bankruptcy proceeding in Switzerland is published in the SHAB. The archive contains millions of entries going back decades. It is the definitive record of Swiss corporate life.
FINMA publications. FINMA publishes circulars, supervisory notices, enforcement actions, annual reports, and guidance documents. All publicly available. All relevant to every financial institution in Switzerland.
ESTV (Federal Tax Administration). Tax statistics, rulings, circulars, double taxation agreements, and administrative guidelines. Published openly.
Curia Vista. The parliamentary database tracking every motion, interpellation, postulate, and initiative in the Federal Assembly. This is the legislative pipeline: what laws are coming, what the debate looks like, and how the political landscape is shaping regulatory outcomes.
Opendata.swiss. The federal open data portal aggregates datasets from across government. Over 10,000 datasets covering demographics, economics, geography, transport, environment, and administration.
Why Nobody Is Using It
The data exists. It is free. It is authoritative. So why is it sitting unused?
Fragmentation. Federal data is in one place. Cantonal data is in 26 different places with 26 different formats. Court decisions are on separate portals. FINMA publications are on FINMA’s site. SHAB is on its own platform. There is no unified access point. A professional who needs to cross-reference a federal law with cantonal implementations and relevant court decisions must navigate multiple systems, each with its own search interface and data structure.
Format inconsistency. Fedlex provides beautifully structured XML. Some cantonal databases provide structured data. Others provide PDF scans. Court decisions range from structured XML to plain HTML to unsearchable PDFs. Building a comprehensive system requires handling all these formats.
Volume. The total corpus of Swiss legal and regulatory data runs into millions of documents. Processing, structuring, and indexing this volume requires significant computational resources and engineering effort. It is not a weekend project.
No AI layer. The data is published for human consumption. Government portals provide search interfaces designed for people who know what they are looking for. They do not provide semantic search, natural language querying, cross-reference analysis, or the kind of intelligent retrieval that AI enables. A lawyer can find a specific article if they know the SR number. They cannot ask “what are my obligations when terminating a commercial lease in Zurich?” and get a comprehensive, multi-source answer.
Perceived as boring infrastructure. AI startups prefer building consumer-facing chatbots or enterprise copilots. The unglamorous work of ingesting, structuring, and maintaining government data at scale does not attract venture capital attention. It is the data plumbing that nobody wants to do.
The Opportunity
Here is what becomes possible when you treat Switzerland’s open government data as a unified, structured, AI-ready knowledge base:
Comprehensive legal research. Every federal law, every cantonal law, every published court decision, linked through citation graphs. Ask a legal question and get answers grounded in the complete corpus, not just the subset that a single portal covers.
Regulatory change detection. Monitor all legislative and regulatory sources continuously. Detect changes as they happen. Alert professionals when provisions affecting their work are modified. No more discovering an amendment after it has already affected a client.
Cross-domain intelligence. Connect legal data with tax data, regulatory data, and commercial registry data. A company change published in the SHAB can be automatically cross-referenced with applicable regulatory obligations. A new FINMA circular can be mapped against the legal provisions it implements.
Historical analysis. With decades of court decisions and legislative amendments, the data supports sophisticated trend analysis. How has the Federal Supreme Court’s interpretation of a particular provision evolved over time? How frequently are certain types of cases appealed? What is the success rate for specific legal arguments?
Multilingual access. Build a system that understands queries in any language and retrieves from all language versions. A French-speaking lawyer in Geneva gets results from German-language BGer decisions that are relevant to their case but that they might never have found searching in French alone.
What It Takes to Build This
Transforming Switzerland’s open government data into a usable AI infrastructure is not trivial. It requires:
Comprehensive ingestion. Every source must be identified, connected to, and monitored. Federal sources via Fedlex APIs. Cantonal sources via individual scrapers or APIs where available. Court decisions from each court’s portal. FINMA publications from FINMA’s website. SHAB from the commercial gazette. Each source has its own update frequency, data format, and access method.
Structural parsing. Raw data must be parsed into structured units. Laws must be broken into articles, paragraphs, and subparagraphs with hierarchical relationships preserved. Court decisions must be parsed into metadata (court, date, case number, legal provisions cited) and content. SHAB entries must be parsed into structured records with entity references.
Citation graph construction. Every reference from one document to another must be identified, extracted, and stored as an edge in a citation graph. A law article that references another law. A court decision that cites a law article. A FINMA circular that implements a legal provision. These relationships are what turn a collection of documents into a knowledge graph.
Vectorization and indexing. Every structured unit must be converted into vector embeddings for semantic search and indexed with BM25 for keyword search. For the Swiss corpus, this means millions of embeddings across multiple languages.
Continuous maintenance. The system must stay current. Nightly syncs with all sources. Change detection for amendments, new decisions, and new publications. Version control so users can see what changed and when.
This is a significant engineering effort. It is also a significant moat. Once built, the comprehensive Swiss legal and regulatory knowledge graph is extremely difficult to replicate. The data is open, but the structured, indexed, AI-ready version of it is not.
Who Should Care
Law firms that want research capabilities beyond what Swisslex and existing tools provide. AI-powered semantic search across the full Swiss legal corpus, with citation graphs and multilingual support, is a qualitative leap in research capability.
Compliance teams that need to monitor regulatory changes across multiple domains. Automated change detection and gap analysis against existing policies.
Financial institutions that need comprehensive, current regulatory intelligence. FINMA circulars, banking legislation, AML requirements, and cross-border obligations in a single, searchable system.
Government and public sector organizations that want to make their own data more accessible and useful to constituents.
Legal publishers that want to enhance their products with AI capabilities without building the infrastructure from scratch.
The First-Mover Advantage
The opportunity is clear, the data is available, and the technology to process it exists. What has been missing is the commitment to do the unglamorous work of building the infrastructure.
The company that builds the comprehensive, AI-ready Swiss legal and regulatory knowledge base first will have a structural advantage that is nearly impossible to overcome. The data is open, but the years of engineering effort, the millions of processed documents, the citation graph with millions of edges, and the continuously maintained pipeline cannot be replicated overnight.
Mont Virtua has built this infrastructure. Our database covers 27,795 federal and cantonal laws, over a million court decisions from 115 courts, 1.4 million citation edges, 2.5 million SHAB entries, and comprehensive FINMA regulatory data. All structured, all indexed, all continuously updated. All delivered through Enclava, our platform for regulated AI intelligence. Visit enclava.ch to access Switzerland’s open government data the way it should have always been accessible.