CASUS is a Swiss legal AI platform that automatically summarises contracts, identifies risks, and answers questions about documents - directly in Microsoft Word or in the browser. All data stays in Switzerland or the EU, with no transfer to the US and no third-party review of content (Zero Data Retention).
Why reading long contracts manually is a real problem
Anyone who reviews contracts regularly knows the situation: a supply agreement arrives, 48 pages, annexes not yet counted. The other side wants feedback by Friday. By the time the contract has been read, the relevant clauses found, and a short assessment sent to a colleague, one to two hours have passed - for a document that may not even go into further negotiation.
According to an analysis by V7 Labs, manually summarising 200 contracts per month generates roughly 400 hours of pure reading effort. That is not an edge case - it is everyday life in M&A teams, compliance departments, and larger law firms.
The goal of AI summarisation is not to replace lawyers. It is to significantly compress the first two hours of a contract review, so that the actual legal work can start sooner.
What does "automatically summarise a contract" actually mean?
AI-powered contract summarisation is not simply shortening a document by half. A well-configured system extracts the legally relevant content: contracting parties, term, liability caps, notice periods, payment terms, governing law - and outputs these in a structured format.
The quality of the output depends on how the model is trained and whether it recognises the contract type. An SPA has different key points than a maintenance agreement or a DPA under Art. 28 GDPR. Generic summarisation tools treat both the same. Specialised legal AI platforms distinguish between them.
A second difference: good tools link the summary back to the specific passages in the original document. Anyone wanting to know exactly where the liability limitation sits does not need to search - they can jump directly to the passage.
How does contract summarisation work with CASUS?
CASUS summarises contracts through the AI Chat - either on demand or in response to targeted questions. Those who want an overview of all key clauses ask for one. Those who want to know what the contract says about termination ask precisely that.
Every answer is linked to the relevant passage in the document. This means the output can be verified immediately rather than accepted blindly. That source connection matters legally, because it keeps responsibility for the assessment with the lawyer - the AI shows, the human judges.
In what CASUS calls Agent Mode, the platform goes a step further: it does not just analyse, it executes changes directly in the document. Insert clauses, adjust wording, check consistency across the whole document - all with correct formatting and numbering, without copy-paste.
After summarisation, a structured risk analysis is one step away. Switching to the Risk and Quality Review lets CASUS identify risks and red flags, assign them to contracting parties, and prioritise by severity (low / medium / high).
This is especially relevant right now, as many Swiss companies are reviewing their contract portfolios for compliance with the revDSG - Art. 30 and Art. 31 revDSG govern the requirements for data processing agreements - and need rapid orientation on content and gaps.
Anyone who wants to see how quickly a 50-page contract can be structured using CASUS can try it directly: Start free at app.getcasus.com/signup.
Which contract types benefit most?
Not every contract is equally suited to automatic summarisation. The biggest time savings come from long, structured documents with recognisable clause patterns.
Share purchase agreements (SPA): An SPA typically contains representations and warranties, closing conditions, MAC clauses, and liability caps. AI summarisation can extract all these points in seconds - a significant time saver for a Zurich-based M&A boutique managing multiple transactions simultaneously.
Service and supply agreements: Term, termination, SLA, and liability caps are the critical points here. Organisations receiving 20 or more such contracts per month recover meaningful capacity through automatic summarisation.
DPAs and data processing agreements: Since the revDSG came into force on 1 September 2023, many in-house teams are continuously checking DPAs for compliance. An AI that shows in seconds whether a contract contains a deletion obligation, sub-processor clause, or return provision substantially reduces manual effort.
For very short, atypical, or heavily individualised contracts, the benefit of automatic summarisation is smaller - these require more manual context.
What AI does not do when summarising contracts
Automatic summarisation has limits that practitioners need to keep in mind.
An AI summary reflects what the document says - but it does not assess whether what was agreed is valid under Swiss law. A liability exclusion clause can be accurately captured and still be unenforceable under Art. 100 OR. That is a legal judgement, not a text analysis.
Equally, a summary typically does not identify what is missing. If a contract contains no governing law clause, a pure summary may not flag that omission. Catching gaps requires a benchmark comparison - checking the document against a reference standard, which is what the CASUS Benchmark workflow does.
And AI summaries are only as good as the quality of the source document. Scanned PDFs with poor OCR quality, handwritten additions, or nested references to external annexes can affect the accuracy of the output.
Practical implications for law firms and legal teams
An in-house counsel at a Basel pharmaceutical company who regularly receives external supply contracts knows: the first step is always the same - get an overview. Automatic contract summarisation removes that step as a manual task.
This changes the workflow, not the responsibility. Whether a contract is acceptable remains the lawyer's call. But the basis for that call is available much faster.
For larger document volumes - such as due diligence involving hundreds of contracts - the CASUS AI Data Room is the right tool. It extracts defined information fields from many documents in parallel and outputs results in tabular form, ideal for Excel-based analysis or reporting.
After analysis, anyone wanting to confirm that a contract is also formally and linguistically correct can run the Proofread module - covering spelling, consistency, and structural issues without altering the legal meaning.
Try CASUS
CASUS is a Swiss legal AI platform, hosted in Switzerland and the EU, with no data transfer to the US. Anyone wanting to review contracts faster, analyse them in a structured way, and work on them directly in Word can try the platform free of charge: app.getcasus.com/signup.
FAQ
What does "automatically summarise a contract with AI" mean?
AI-powered contract summarisation extracts the legally relevant key points from a contract - parties, term, liability, termination, governing law - and presents them in a structured format, without requiring the document to be read in full manually.
How quickly does AI summarise a 50-page contract?
With a specialised tool like the CASUS AI Chat, summarising a 50-page contract typically takes a few seconds to under a minute. Manual first reading of the same document usually takes one to two hours.
Is an AI summary legally reliable?
An AI summary reflects what is written in the document - source-based and traceable. It does not replace a legal assessment of whether a clause is valid under Swiss law (for example under Art. 100 OR). The legal judgement remains with the lawyer.
Which contract types are best suited to automatic summarisation?
Longer, structured documents work best: SPAs, NDAs, supply agreements, service contracts, and DPAs. Very short or heavily individualised contracts without standard clause patterns benefit less.
Where does my data go when using CASUS?
CASUS hosts all data in Switzerland and the EU. There is no data transfer to the US. CASUS operates with Zero Data Retention and no manual review by third parties.
Can I make changes to the contract directly after summarisation?
Yes. CASUS Agent Mode executes changes directly in the Word document - inserting clauses, rewriting text, checking consistency - with correct formatting and numbering applied automatically.
What is the difference between summarisation and benchmarking?
A summary shows what is in the contract. A benchmark comparison shows what is missing or deviates from a standard - measured against a reference (such as an internal playbook or best practices for NDA or SPA). CASUS offers both functions.
Can CASUS summarise many contracts at the same time?
For parallel analysis of many documents, the CASUS AI Data Room is the right tool. Dozens or hundreds of contracts can be uploaded simultaneously and extracted according to defined fields - as tabular output, suitable for due diligence or compliance reporting.
