CASUS is a Swiss legal AI platform that automatically reviews framework agreements for risks, missing clauses, and deviations from your internal standard. The Risk & Quality Review identifies findings by severity and assigns them to the relevant contract parties. Data stays in Switzerland or the EU - no transfer to the US, no human review, zero data retention.
Why framework agreements deserve extra scrutiny
Framework agreements govern business relationships that can span years: supply conditions, liability limits, price adjustment mechanisms, termination rights. A single imprecise paragraph can carry material financial or legal consequences across hundreds of individual call-offs.
In practice, framework agreements are often reviewed less rigorously than one-off contracts. The reasoning goes that they are familiar documents, already in use. That familiarity is exactly what makes them risky.
Research from German mid-market companies shows that employees in contract-heavy organisations spend between 8 and 12 hours per week on contract-related tasks - around 400 to 600 hours per year. A significant share of that time goes into reviewing and tracking existing framework documents.
AI-assisted contract review addresses this directly. It brings consistency, speed, and structured documentation of review results - without replacing the lawyer's legal judgment.
What makes a framework agreement different from other contract types?
A framework agreement sets the baseline terms for many future individual transactions without itself constituting those transactions. It is not a purchase agreement or a service contract in the conventional sense; it is a set of rules that activates on demand.
This has a concrete implication for the review: it is not enough to evaluate individual clauses in isolation. Reviewing a framework agreement means understanding how the contract structure works as a whole - for example, how a liability clause interacts with a price adjustment rule when a specific order is triggered.
That is precisely where linear, manual review processes fall short.
5 things AI spots in a framework agreement immediately
1. Missing or incomplete liability clauses
Liability without a cap is a classic red flag in framework agreements. AI does not just detect the absence of a liability limit - it also picks up incomplete drafting, for instance where an exclusion covers direct loss but says nothing about consequential damage.
The CASUS Risk & Quality Review prioritises findings by severity (low / medium / high) and provides concrete drafting options for each issue, which can be applied directly in Word without copy-paste.
2. Gaps in data protection and confidentiality
Framework agreements often include a confidentiality clause but no explicit rules on data deletion or data processing arrangements. For contracts subject to the Swiss DSG or the EU GDPR, that is a genuine weak point.
The Benchmark workflow checks the document against a defined standard - an internal playbook or established best practices - and flags missing data protection topics as gaps. The match is shown as a percentage score, making it immediately clear how far the contract deviates from the required standard.
3. Unclear or missing termination provisions
Many framework agreements include a term clause but no clear rule on extraordinary termination or on what happens to open call-offs if the agreement ends. That is a recurring source of disputes.
AI detects when termination notice periods are contradictory, when a right to terminate for cause is absent, or when the clause is present but drafted one-sidedly in favour of one party. The AI Chat can be asked directly how termination is handled in the document - the answer links to the relevant passage, so it is easy to jump straight to the source text.
4. Price adjustment clauses without clear mechanisms
Price adjustment clauses in framework agreements are often vague: "prices may be adjusted annually" says little about which index applies, within what period, and with how much notice. That is a typical negotiation risk that AI flags reliably.
CASUS AI Chat can evaluate exactly what price adjustment logic applies based on the contract text. Agent Mode can then insert a more precise clause directly into the document at the right place, with correct numbering and formatting.
5. Inconsistent definitions and cross-references
In long framework agreements - especially those that have been amended over the years - inconsistencies accumulate. A term is defined in clause 2.1 but spelt differently in clause 8.4. A cross-reference points to a section that no longer exists.
The Proofread module checks exactly these formal consistency issues: definitions, cross-references, annexes, placeholders, and contradictions. It is not a substitute for a legal assessment, but it is a reliable final check before the document goes out.
How AI concretely supports framework agreement review
AI-assisted review is not a one-size-fits-all process. Different workflows make sense depending on the situation.
When reviewing a new framework agreement for the first time, the natural starting point is the Risk & Quality Review. The output is a structured list of findings with severity ratings and drafting suggestions - a solid basis for negotiation.
When the goal is to enforce an internal standard - for example, when a supplier has submitted their own framework agreement - the Benchmark workflow is the right tool. It compares the third-party document against an internal playbook and makes deviations systematically visible.
When a dozen or more framework agreements need to be analysed in parallel - during due diligence or a supplier audit - the AI Data Room handles that. It extracts defined fields from all documents into a tabular overview: contract duration, liability caps, notice periods, governing law, and more.
When the legal classification of a specific clause is unclear, the Legal Research mode provides source-based assessments drawing on more than 660,000 cantonal and federal court decisions as well as statutory law.
CASUS runs directly in Microsoft Word or as a web app. All data is hosted in Switzerland or the EU - no transfer to the US, no human review, zero data retention.
Legal teams that review framework agreements regularly and need structured, traceable results can try CASUS free at app.getcasus.com/signup.
Practical tips for using AI on framework agreements
A few points from practice that make a real difference:
Provide context. AI review tools work better when the contracting party is known. CASUS identifies parties automatically and analyses risks from each party's perspective - which is directly more useful than a generic list of clause risks.
Define a standard. Teams without an internal playbook leave value on the table. Even a basic clause template for the relevant contract type is enough to make the Benchmark workflow meaningful.
Treat AI as the first layer, not the only one. AI-assisted review is not a substitute for legal judgment - it is a systematic first pass that reduces the risk of missing something. Assessing the legal position and negotiating the contract remains the lawyer's responsibility.
Document the findings. Structured findings from the review workflow can serve as the foundation for a negotiation strategy - and as an internal record that a review took place.
FAQ
What is a framework agreement and why does it need careful review?
A framework agreement sets the baseline terms for many future individual transactions. Because it typically applies over years and across many call-offs, errors or gaps have long-term consequences, which makes thorough review particularly important.
Which clauses are most commonly problematic in framework agreements?
Common weak points include liability clauses without a cap, vague price adjustment mechanisms, missing data protection provisions (particularly under the Swiss DSG or GDPR), one-sided termination rights, and inconsistent use of defined terms throughout the document.
How does AI review a framework agreement in practice?
AI tools like CASUS analyse the contract text in a structured way: they identify the parties, classify clauses by topic, compare them against a defined standard, and prioritise deviations by severity. Results are presented as structured findings with concrete improvement suggestions.
Can AI replace a lawyer when reviewing a framework agreement?
No. AI supports the review through fast, consistent analysis and structured outputs. Legal assessment, negotiation strategy, and responsibility for the final contract remain with the lawyer.
How secure are contract documents when using AI tools?
This depends heavily on the provider. CASUS hosts all data in Switzerland or the EU, does not transfer data to the US, operates without human review, and applies zero data retention. Details are available on the CASUS security page.
What is the difference between Risk Review and Benchmark for framework agreements?
The Risk & Quality Review analyses the contract for risks and weaknesses from the perspective of a specific party. The Benchmark compares the contract against an internal standard or best-practice playbook and shows which clauses are missing or deviating. The two workflows complement each other.
Is AI also suitable for analysing many framework agreements at once?
Yes. The CASUS AI Data Room supports parallel upload and structured extraction from large document sets. This is well suited for due diligence processes, supplier audits, or reviewing an existing contract portfolio.
Does AI review work directly inside Microsoft Word?
CASUS runs as a Word add-in, so the review takes place within the familiar working document. Improvement suggestions can be applied directly without copy-paste, correctly formatted and placed at the right position in the document.
