Zürcher Kantonalbank: A Pioneer Bank in the AI Era

Zürcher Kantonalbank (ZKB) is one of the largest cantonal banks in Switzerland and one of the country's most important financial institutions. As one of the very first banks to use artificial intelligence in legal contract review, ZKB sets a clear signal: legal & compliance has to embrace digital tools — not despite the high regulatory bar in banking, but because of it.

When the legal team recognised the need to free capacity for strategic, business-critical work, they opted for AI support to standardise high-volume, low-differentiation contract reviews and shift them to procurement.

At a glance:

• One of the largest cantonal banks in Switzerland, founded in 1870
• One of the first banks worldwide to use Legal AI in contract review
• Use case launch: NDAs and Data Processing Agreements (DPAs)
• Reviews shifted from specialised Legal Counsel to Procurement specialists

The Challenge: High Volume, High Compliance, Few Lawyers

The legal department cannot personally touch every contract in a heavily regulated bank. Quality and compliance standards are non-negotiable — yet capacity is finite. The business has to be able to handle standardised contracts itself, without compromising the bar that financial regulators demand.

1. Capacity bottleneck: Specialised legal counsels were tied up reviewing high-volume, low-differentiation contracts (NDAs, DPAs) that did not require their full expertise.

2. The standardisation gap: Manual reviews depended on individual interpretation. Reaching uniform quality required intensive training and continuous controls.

3. Compliance pressure: Banks operate under constant cost and efficiency pressure, while compliance requirements remain particularly high. There is no room for shortcuts.

4. Cultural barrier: Most users had never worked with comparable AI software. Trust in automated review processes had to be built carefully — and continuously.

5. Strategic overload: When Legal Counsel are busy with routine reviews, the capacity is missing for the truly business-critical, strategic legal questions.

The Solution: AI-Powered NDA & DPA Review in Three Steps

Together with Legartis, ZKB rolled out a structured AI-supported review process for NDAs and DPAs — designed so that procurement specialists without legal training can run reviews independently, while company-specific rules are enforced automatically.

Step 1 – Needs Analysis & Process Design
Before evaluating any tools, ZKB analysed needs, working methods and processes. LegalTech was selected to fit existing workflows — not the other way around.

Step 2 – Playbook Standardisation
Company-specific requirements for NDAs and DPAs were captured in playbooks. What previously required intensive training and manual controls now happens automatically — with consistent quality across every review.

Step 3 – Self-Service Review by Procurement
Procurement specialists run NDA and DPA reviews independently, with playbook guidance. Legal Counsel step in only for cases that fall outside the playbook — for example, when negotiations cannot stay within defined parameters.

Results: What Has Really Changed

Standardised review for NDAs and DPAs
Two high-volume contract types now run through an automated, consistent process. Quality is enforced via the playbook — not via individual interpretation.

Procurement empowered to review autonomously
Specialists without legal training handle contract reviews at the same quality level as before — no quality loss, no compromise on compliance.

Legal Counsel freed up for strategic work
The legal team focuses on truly business-critical questions and supports procurement only for negotiations outside playbook parameters.

First-mover position among banks
ZKB is one of the very first banks to use AI for legal contract review — sending a clear signal to the financial industry about the role of LegalTech.

Cultural shift toward digital legal
Pilot users moved quickly from initial scepticism to trust. The transition from "I have to read every contract" to "the AI has already done that" is real — and ongoing.

Foundation for further use cases
NDAs and DPAs were the deliberate starting point. Licence agreements and industry-specific contract types are next.

"Banks operate under constant pressure to simplify processes, become more efficient and reduce costs — while compliance requirements remain particularly high. Recognising and using the potential of LegalTech and AI is therefore a duty of any legal service."

— Dr. Tom Fischer, General Counsel, Zürcher Kantonalbank

Implementation Learnings: What Tom Fischer Tells Other General Counsels

Start with the process, not the tool
ZKB analysed needs, methods and processes first — and only then evaluated LegalTech solutions. Tools that fit existing workflows succeed; tools that don't, sit unused.

Cultural change has to be guided
AI is new for most users. Trust in automated reviews builds over time, not on day one. Continuous accompaniment matters more than the rollout itself.

Pick a partner with real proximity
For ZKB the deciding factors were Legartis' reputation in the LegalTech industry, an intuitive interface, Zurich-based developers and engineers, and flexibility in covering ZKB's specific needs.

Begin where the value is highest, not the risk
NDAs and DPAs are high-volume, low-differentiation contract types. Perfect to build first experience and trust before extending to more complex contracts.

The Evolution of Legal AI at ZKB: Where the Bank Stands — and Where the Journey Is Heading

Stage 1 – Paper files and manual review (past)
Until recently, parts of Legal & Compliance still ran on paper files. Reviews depended entirely on individual lawyers.

Stage 2 – AI-supported standard contract review (today)
NDAs and DPAs are reviewed by procurement with playbook-guided AI. Legal Counsel takes over only for exceptions and strategy.

Stage 3 – Extended scope and industry-specific contracts (next)
Licence agreements and bank-specific contract types come next — developed jointly with Legartis on top of the same foundation.