Legal research is one of the most time-intensive tasks in legal practice. Searching Federal Supreme Court decisions, reviewing cantonal case law, locating the right statutory provisions - it takes hours that are needed elsewhere. AI-powered research tools promise real relief. But which available solutions genuinely fit the Swiss legal environment, and where do the limitations begin?
This article gives a structured overview of the most relevant research tools for Swiss law firms and in-house legal teams - looking at data protection, language capability, source coverage, and practical usability.
Why international tools often fall short
The market for legal AI platforms is expanding fast. Harvey, Legora, and Spellbook are well-known names - but none of these tools were built primarily for the Swiss market. That shows up in concrete ways.
US platforms are subject to the CLOUD Act. Even if data is stored on European servers, US authorities can, under certain conditions, demand access. For Swiss lawyers bound by professional secrecy under Art. 321 StGB, this is not an abstract theoretical risk.
Then there is the question of training data. Many models were trained predominantly on English-language legal texts from US or UK law. Swiss norms such as OR, ZGB, or ZPO appear at most on the margins. Using these models to research Swiss legal questions risks structurally weak outputs - even when the text reads fluently.
And many tools lack direct integration with Microsoft Word. Anything not integrated creates workflow friction: results have to be copied, reformatted, and manually inserted. That costs time that the automation was supposed to save.
What matters when evaluating research tools for the Swiss market
Law firms and in-house teams evaluating a research tool should check four aspects systematically:
Data hosting and professional secrecy: Where are queries processed? Is there zero data retention? Does no human review take place? Tools that store queries or allow third-party review are difficult to reconcile with professional secrecy obligations.
Source coverage: Does the database include Federal Supreme Court decisions, cantonal case law, and current statutory texts? Are sources cited transparently?
Language capability: Swiss law firms work in German, French, and Italian - sometimes within the same matter. A tool that functions well only in English rules itself out for many workflows.
Workflow integration: Can the tool be embedded in existing processes, particularly Microsoft Word?
The main tools at a glance
DeepLaw
DeepLaw is a Swiss-built platform specialising in access to Swiss legal sources. Its database covers federal legislation (continuously synchronised via Fedlex), cantonal legislation (via Lexfind, in each canton's official language), and court decisions going back to 1954. Decisions are automatically translated into German, French, Italian, and English.
On the data protection side, DeepLaw holds ISO 27001:2022 certification, processes queries exclusively on Swiss servers, and states that no user-identifying logs are retained.
For firms whose primary need is source retrieval - finding statutes and decisions - DeepLaw is a well-positioned option.
Lexi Search
Lexi Search enables natural-language search across Swiss court decisions without requiring exact keyword matching. The platform covers Federal Supreme Court decisions and cantonal case law, offers integrated translations, and can be connected to existing AI workflows via open MCP standards. Pricing is based on a flat rate rather than per-query billing.
Lexi Search is aimed at legal professionals who want to get started with case law research quickly, without complex setup.
Lexplorer
Lexplorer is an API- and MCP-based research infrastructure for Swiss legal sources. The platform provides programmatic access to over 660,000 federal and cantonal decisions, statutory articles, and legal literature — combining vector and semantic search with zero data retention on Swiss infrastructure.
Lexplorer is primarily aimed at developers and organisations looking to embed legal research into their own applications, knowledge management systems, or client portals. CASUS currently works with Lexplorer as the data infrastructure powering its Legal Research feature.
Omnilex
Omnilex is a Zurich-based legal AI platform with a proprietary multi-agent framework for source-cited answers to complex questions under Swiss law. The platform includes over 200,000 AI-generated legal annotations, document upload for proprietary files, and multiple modules — including a Word add-in, Playbooks, Grid, and Cases. Omnilex is now in use at over 120 organisations across the DACH region.
For firms and in-house teams that need commentary functions and structured document analysis alongside research, Omnilex is a fully developed offering.
CASUS: research inside the working process
CASUS, a Swiss legal AI platform, is available as a Microsoft Word add-in and a web app. Hosting is in Switzerland and the EU; no data is transferred to the US. Zero data retention and no human review are configurable operating parameters, not marketing claims.
The Legal Research mode is embedded directly in the AI chat. It provides access to over 660,000 federal and cantonal decisions as well as statutory articles. In practice, this means relevant reasoning sections (Erwagungen) are highlighted directly in search results, and a preview of decision content appears inline within the answer - without needing to open the full text separately.
Outputs are source-based, structured, and traceable. CASUS does not return generic web content; it produces risk analyses of legal positions with risk drivers, pro and con argument lines, and concrete recommendations - for example, whether to hold a position, offer a fallback, or add a clause or disclaimer.
What sets CASUS apart from pure search tools is that research results feed directly into further work. Via chat actions, clauses from the research can be inserted into a document, rationales for internal assessments can be drafted, or texts can be adjusted - without switching tools, directly in Word.
More detail is available on the Legal Research product page.
Research as part of a broader workflow
Legal research is rarely an isolated step. Reviewing a contract usually requires both: a look at the relevant case law and an analysis of the document at hand.
CASUS connects both in one platform. The Risk & Quality Review identifies risks and weaknesses directly in the contract; the Benchmark workflow checks the document against internal playbooks or industry standards. The AI Proofread ensures linguistic and formal consistency before sending.
For larger review tasks - in due diligence processes, for example - the AI Data Room allows parallel analysis of dozens or hundreds of documents with structured tabular output.
Firms that want to consolidate research, document analysis, and drafting in a single tool will find that CASUS was built for the Swiss context.
What general AI tools do - and where they stop
ChatGPT and comparable models are used daily by lawyers to structure text, formulate arguments, or get initial orientations. That is legitimate and saves time.
But general language models have no connection to current Swiss case law. They cannot cite Federal Supreme Court decisions from the past year because they do not have access to them. Sources are neither disclosed nor verifiable. Building legal positions on that basis means working on an uncertain foundation.
Useful for orientation - not suitable for substantive legal research.
Try CASUS
Swiss law firms and in-house legal teams that want to test structured legal research with direct Word integration can start with CASUS at no cost. All data stays in Switzerland or the EU, with no transfer to the US.
Start free trial - or learn more about data security and hosting.
FAQ
Which Swiss legal sources does CASUS cover?
CASUS provides access to over 660,000 federal and cantonal decisions as well as statutory articles. Relevant reasoning sections are highlighted directly in search results, and a content preview of decisions appears inline within the answer.
Is CASUS compatible with Swiss professional secrecy obligations?
CASUS processes data exclusively in Switzerland and the EU. No data is transferred to the US. Zero data retention and no human review are configurable operating parameters, not marketing promises.
What distinguishes specialised legal AI tools from general language models like ChatGPT?
Specialised legal AI tools connect to current legal databases and cite sources transparently. General models like ChatGPT have no access to current Swiss case law and cannot provide verifiable source references.
Does CASUS work directly in Microsoft Word?
Yes. CASUS is available as a Microsoft Word add-in. Research outputs, drafting suggestions, and document changes can be applied directly in the document without switching tools.
What does the CASUS Legal Research mode actually produce?
The Legal Research mode produces structured risk analyses of legal positions with risk drivers, pro and con argument lines, and actionable recommendations. Outputs can be reused directly for internal assessments, clause rationales, or text adjustments via chat actions.
Can CASUS be used for due diligence research?
The CASUS AI Data Room allows parallel analysis of dozens or hundreds of documents. Content is output in structured tables, which suits clause matrices, compliance checks, and due diligence processes.
Which languages does CASUS support?
CASUS is designed for the multilingual Swiss legal environment and supports German, French, English, and Italian.
What is the difference between CASUS, DeepLaw, and Omnilex?
DeepLaw is a specialised search solution for Swiss legal sources with broad source coverage. Omnilex provides AI-generated legal commentary and modular document analysis. CASUS combines legal research with contract analysis, document editing, and Word integration in one platform — keeping research and the working process directly connected, without switching tools.
