Most legal departments grow organically. They may start with one internal lawyer, then add another, and little by little each team member owns different projects and focuses. This approach works for some settings; the challenges arrive when more and more work is added to the legal team. A high-performing legal department is built with this in mind—it’s designed for the long run, built to last and work efficiently over time.
This is exactly what Jennifer Graf, General Counsel, Legal and Organizational Consultant, and Founder of legalmeetscontext, and Dr. Nadine Lilienthal (Head of Legal Expertise and Alliances DACH, DiliTrust) discussed in a recent webinar. They unveiled how to build a high-performing legal department from scratch, putting humans first and letting the tech play its role when the time is right.
Graf’s experience comes from a German Hidden Champion—a billion-euro measurement and data technology company with over 8,000 employees that hired its first in-house lawyer after more than 100 years in business.
Why the Classic Corporate Approach Is Limiting
Graf recounts her experience building a high-performing legal department from the ground up. After years of working as an in-house lawyer in a well-structured large corporation, she made a bold move to the mid-market and was tasked with building the entire legal function. Today, her team consists of seven people: five fully qualified lawyers and two colleagues who handle organizational matters and manage the department’s software tools and management systems.
Her first instinct was to follow the corporate playbook: divide responsibilities by subsidiaries, assign topic owners, and implement a ticketing system. Within a month, the approach stopped working.
I remember a team meeting where things got really heated. Different colleagues were doing the same work for different stakeholders but in completely different ways. Nobody was aligned.
At first, one can be tempted to blame people’s behavior, but a principle in change management offers a different perspective: “conditions create behavior.” Rather than fixing people’s behavior, it’s worth understanding the framework and environment around them, fixing that first, then attempting the change.
The business-first approach
It may seem counterintuitive at first, but starting from the business rather than the law first is beneficial when it comes to legal department transformation. It begins by taking a radical step back and asking simple questions: What kind of company are we actually working in? What does the legal function need to look like and what is it expected to do?
Step 1: Building the Company Portrait
The company portrait in plain language is simple to make, yet for legal professionals going through this step may not be so obvious. For Graf and her team, this meant laying out a simple description, not in legal jargon, of what the company does. The description was:
- A company headquartered in Germany
- A high-precision measuring instruments manufacturer, a niche player
- Exclusively B2B—no consumer business
- Owned by a charitable foundation, meaning it cannot become public
- Reinvests profits into R&D
- Still produces entirely in Germany
Getting to that plain-language company portrait was actually the hardest part, because lawyers tend to stay in their own way of thinking.
Jenifer Graf
With this crucial information, the team could move forward and focus on their role within this structure: What risks could they encounter in this business setting?
Step 2: Mapping the Business Risks
To better understand the risks and each team member’s role, the process is simple. Everyone described their daily tasks and core concerns without categorizing or templatizing them. This is what changed the game for Graf’s team: no categorization, no templates—just honest descriptions. The team clustered the results and compared them against the company’s portrait.
Step 3: Creating the Legal Department Map
What emerged was a “Legal Department Map”, a sort of visual orientation tool showing:
- Core topics: what the company truly pays the legal department to handle, the topics that truly require full intervention and expertise
- Peripheral topics: important topics that are never a priority and that can potentially benefit from automation or standardization
- Foreign topics: tasks the legal team was doing but that actually belonged somewhere else entirely
For Graf’s team, R&D and IP was not a core topic, nearly all development happened internally with standardized templates. On the other hand, immovable property became core due to strategic land acquisitions for wind turbines and geothermal energy, making legal central to the company’s survival strategy.
Paving the way for legal AI
Implementing new tech only works under certain conditions. Among them are a true understanding of the tech by end users and a real solution to specific pain points. With the AI frenzy, there is no difference, except that under pressure, teams sometimes want to hurry the process, and it can end badly.
I see it often starts with the question: which tool should we use? From my perspective, the right question is: what is the use case that we want to tackle? Where do we want to deploy AI?
— Nadine Lilienthal
Graf’s case is the perfect example of this. Once the priorities and roles were clear, the team was ready to deploy tech solutions strategically, not reactively. This intentional approach to AI implementation ensures technology serves the department’s mapped priorities rather than creating new complexity.
Differentiating Probabilistic vs. Deterministic AI
It is crucial to understand the difference between AI types, because this builds a better foundation for assessing the tech that best fits your needs. Lilienthal explains a key distinction: the one between probabilistic and deterministic AI.
- Probabilistic AI covers popular LLM models such as Claude, ChatGPT, Legora, or Copilot. It works based on probabilities, meaning the answer is not 100% correct but always the most probable answer. This AI type works best for senior and experienced legal professionals who need fast outputs but have the capacity to evaluate the results and make corrections.
- Deterministic AI covers if-then logic models such as template generators. This AI model produces the same output if given the same input, and is 100% correct, therefore. This AI type is best for empowering non-legal or less experienced legal professionals with controlled results.
Understanding these AI types helps a high-performing legal department map use cases, just like mapping the department itself. This is what choosing AI strategically based on use cases looks like in practice.
Practical Use Case Examples
Lilienthal walked the audience through real scenarios where AI can transform a high-performing legal department’s efficiency:
| Use case | AI type | How it works | Key benefits |
|---|---|---|---|
| NDA review | Probabilistic | Business uses AI tool (e.g., DiliTrust Risk Detector) to flag red flags based on legal’s playbook → escalates only when issues appear | Frees legal from high-volume, low-risk work |
| Template generation | Deterministic | Sales answers a questionnaire → system assembles a contract from a pre-approved clause library → legal maintains full control of it | Update one clause once, then it propagates across all templates automatically |
| Strategic crisis response | Probabilistic | AI analyzes thousands of contracts in hours during geopolitical disruptions to answer: Where are biggest exposures? Which customers to prioritize? | A 90–95% accurate picture in two hours enables fast strategic decisions |
Even if AI makes mistakes in the strategic response scenario, a 90–95% accurate picture in two hours is infinitely better than no picture at all. For strategic decisions, you don’t always need 100% accuracy — you need direction. Nadine Lilienthal
High performance legal departments run on business, law, then tools
Building a high-performing legal department requires starting with the business, not the law. It means mapping priorities before deploying technology, and choosing AI strategically based on use cases and user capabilities. Graf’s journey, from a one-person show to a strategic partner at the executive table, demonstrates that a high-performing legal department evolves from protective shield into strategic navigator when built on the right foundation.
This transformation is echoed by recent findings from Bucerius Law School’s AI-Conomics of Legal Services study. For legal leaders tasked with building or restructuring their teams, the message is clear: understand your business, map your department, then deploy technology with precision. A high-performing legal department is not just efficient. It is strategic, sustainable, and ready to lead.
