In this in-depth conversation between David A. Bloch, CEO of Legartis, Darko Veljača, CRO of Legartis and Alisha Andert, MD & Co-Founder of This is Legal Design, they discuss the approach of Legal Design that builds the bridge between technology and human needs.

They will talk about the qualities of AI, practical applications, challenges, intelligence, business acumen and creativity.

The dialog offers valuable insights into the direction the AI world is moving in - and what this shift means for both innovation and legal integrity.

The importance of Legal Design

Darko: What is Legal Design and why is it so important now?

Alisha: Legal Design is often misunderstood, especially in the legal context, where many people initially think of the term “design” as a purely visual discipline - for example, the visual preparation of contractual documents. Although this is also part of the practice, Legal Design is essentially a much more comprehensive approach to innovation.

The aim of Legal Design is to make legal processes and services more user-friendly and to create real added value for users. A central methodology here is design thinking - a globally established innovation approach that relies on a consistent focus on the needs of users. Solutions are developed, tested and refined iteratively.

Especially in the context of rapid technological development, it is clear that it is not just a matter of using technology at any cost. Rather, it must be integrated in a meaningful way so that it actually creates benefits. This is precisely where Legal Design comes in: to enable better legal processes, more efficient tools and more effective communication in legal services.

New AI business models 2025 in Legal Design

Darko: What new business models will emerge from the combination of AI and Legal Design in 2025?

Alisha: In 2025, the combination of artificial intelligence (AI) and Legal Design will give rise to new business models aimed at greater individualization of legal services. The focus here will be on tailoring services to specific user needs - both in legal departments with internal clients and in law firms with external clients. Services can be increasingly fine-tuned and tailored to personal requirements with the help of AI. For example, employees from the marketing department have different legal needs than those from the HR department; the requirements also differ depending on previous legal knowledge.                                      

One business model results from the use of AI-supported assistants that support legal users with research tasks or answering legal queries in their area of expertise. Other business models are emerging in the area of user-friendly legal communication: Legal Design uses generative AI to simplify complex legal texts such as company guidelines or contract documents so that they are understandable for different target groups - for example, by using prompts such as “explain it to me like I'm five”. This allows engineers or HR managers, for example, to directly identify the section relevant to them.

Another specific field of application is the automation of contract reviews. AI can extract rules from previous review processes and transfer them into playbooks that make it easier to work with contract management tools. This development is already being implemented in practice and is considered a business model with great future potential.    

New opportunities for disruption in the legal department 

Darko: Where do you see the greatest opportunities for disruption in legal departments and where do you think real change is on the horizon?

David: The greatest opportunities for disruption in legal departments lie in numerous use cases “where work can really be done differently in a legal department”. Historically, legal departments are still “perceived as blockers, as cost centers” and are “certainly not the favorite department to talk to”. However, this perception “can change” because “there are many opportunities or use cases where this can be done”.

The central question is “how can we work more efficiently?”, i.e. “do as much or even more with the same amount of resources or potentially even fewer resources”. This is linked to the way “to get away from a legal department, and therefore also a company, operating blindly”. Another lever for disruption is “to design the collaboration model differently internally in a company” and “to operate and interact differently with external customer suppliers”.

There are also opportunities to “not only retain talent, but also attract it”. All of these approaches are possible “using AI” to “really rethink your own work”.

Real change is characterized by the direct and fundamental use of technology: "It's not just about introducing technology ... but really thinking fundamentally about how can I use AI and, in particular, how can I ensure that what I have available can do much more with it?" This is exactly where “the interface between AI and Legal Design” comes into play: “It's not just about working with technology in some way, but it's really about making it better through the combination”, “to provide a solution that is sensibly tailored to a legal department, to a specific person ...”.

The potential of AI in the legal sector 

Darko: Where would you currently invest your time and money?

David: As an AI company, we are currently investing heavily in speeding up the onboarding process. Why? Because we have had precisely this experience. Companies come to us, want to start with legacy, but don't have any company guidelines written down anywhere. That is, if at all in the minds of the individual people, then you form a working group and then you discuss it, but it's a long process and we want to simplify it, we want to speed it up and that's exactly where the latest innovation from Legartis comes in. So we are really introducing a process. You can imagine it like this: I go into a conversation with a chatbot, explain, who am I? I am a tech company. I would like to negotiate a service contract with a branding agency. The following points are important to me. And then I enter two or three negotiated contracts, as well as two or three contracts that may have been commented on first. And then the MKI starts to work.

And what Legartis does there is really firstly, it automatically creates a first draft of a playbook. But that's not all; the individual test points come with the playbook. There is a test set for each checkpoint, because I also have to somehow ensure the quality of the AI, and the test set also automatically provides a so-called F1 score, i.e. a score that shows me how certain the AI is that the result is correct. And that's how we speed up this onboarding process.

Basically, however, it's certainly about AI companies like us moving from systems of individual prom to truly agent-based systems. In other words, we want to ensure that AI can think, can act and can really be integrated into workflows.

And as a client, as a legal department today, I would clearly invest in two things. Firstly, I would make sure that I can cover the low-hanging fruit first, so that I can also celebrate successes quickly. And secondly, I would invest in technology partners who, on the one hand, show that they are really at the forefront, but who can also demonstrate how they deal with difficult issues such as how to ensure the quality of the AI.

Darko: Does the legal AI sector still have some catching up to do in terms of quality?

David: My short answer: Yes. There are providers who are working intensively on this topic and initiatives that are dealing with it. The Legal Tech Association in Germany has launched an initiative on benchmarking. Benchmarking is already very complex because you first have to be aware of what you are actually comparing.

When you talk to providers specifically, it becomes clear that the issue of quality is central in my view. There are two different aspects to this: Firstly, just talking about quality, and secondly, actually taking action and demonstrating what is being done to ensure quality.

Darko: Alisha, are customers now approaching you and your team with completely new requests relating to AI?

Alisha: We totally notice the hype. The requests have changed a little to the extent that they are more about AI than about Legal Tech in general. But what we're already seeing is that it's actually often about core Legal Tech or digitalization issues and they've just been given a slightly different label. Nowadays, not a single department can duck away from this. The pressure is coming from the management, so the CEOs of the companies really need to approach the heads of the individual departments and say, what is your AI strategy? General Counsel have to be very clear about what they are doing, what their plan is and how they want to deal with it in general.

That's why many people want to know which tools they can use. They want to define use cases for themselves and I would also say that most of them are more or less still at this level. Some have already set up test rooms with individual tool providers and that's exactly where we come up against these challenges, namely that once we have selected a tool (hopefully specifically based on certain parameters), then comes the really big part, namely actually putting the whole thing on the road, implementing it and deriving real added value from it. And I think we're currently at an exciting point here, because many people then withdraw relatively quickly, try things out and say that it didn't work for us because they didn't really try it out.

Long story short, we see a total hype, but I still see a bit of difficulty from that initial hype to: How do we actually integrate this into our processes now or how do we fully transform our processes? Lawyers tend to think that something has to work immediately. This mindset that we can also approach a solution with iterations or that this is even a basic requirement in an environment with so many uncertainties such as a technical development, as we are currently experiencing, has not yet arrived and, even if it is understood, has not yet been internalized. It's always a bit like we want to be the jack of all trades and we want it to work immediately.

And that's where many people will first gather their experience and then perhaps make the most important experience, namely to see that you can get closer to the whole thing through iterations and hopefully not chalk it all up to defeat, but say, okay, we've learned this and that from it. It could be that we were looking at the wrong functions. The use case was not the one that brought us the greatest added value. But it could also be that we didn't invest enough in taking people with us, in terms of change management and implementation. So I think it's very important to really extract these learnings. At the moment, it always has to work straight away.

Darko: What changes are taking place in the legal AI sector?

David: Customers know much more specifically what they are looking for today - initial discussions with interested parties and legal departments are much clearer than they used to be and it is possible to identify more precisely what they are looking for. In the past, people used to go “windows shopping”, whereas today it is expected that it really has to work right from the start. At the same time, people are still developing and learning - but the mindset of lawyers still gets in the way a little, for example the “80:20 mindset” - in relation to the Pareto principle - is simply not really there yet“, which creates a ”contradiction".

There is also a lack of experience, for example in “purchasing and introducing software and ”this has to be built up". There should be a willingness, for example, to “really make time and resources available”. At the same time, Legal Tech providers have been able to expand their knowledge in these areas, which is why we can talk about this combination today. We couldn't do that without further development in this area. - “a mirror of what we are also experiencing in the market”.

"AI already offers many, many use cases for legal departments and lawyers. People are still lagging behind in terms of expectations, and Legal Design can step in and, in principle, massively support expectation management and the formulation of use cases. Lawyers should perhaps move away a little from the idea that everything has to work straight away and immediately, but that there is in fact a learning curve, as with all technologies. AI is currently developing at a rapid pace, and my takeaway for myself is that if you stand still now in principle, you really will stand still in time.

Example 

David: Customers started a longer pilot phase a good two years ago. It was simply about a specific use case that was taken at the customer's request so that they could go through a learning phase and see how it works. This took 6 months and created a certain amount of enthusiasm in the customer's legal department. One particular person prevailed internally and convinced with Legartis. Further use cases were then made available to the legal department and the customers were of the opinion that the product was not only sufficient in the legal department, but also needed to be used in the business in order to gain real efficiency from it. Legartis therefore developed the solution internally and ensured that it could be used as a self-service. It was a process of taking time and learning, which suddenly generated so much enthusiasm that they were given resources internally to do even more. “I thought it was a super cool and nice use case.”

Would you like a quick overview of all your active contracts?

Challenges in the use of AI in companies

Darko: What are currently the biggest challenges for the practical implementation of such use cases at companies?

David: 

Alisha: The biggest challenges in the practical implementation of AI use cases in companies lie firstly in change management: the joy of innovation alone is not enough if not all employees are on board. It is therefore crucial to involve those affected early on, to understand their needs and - ideally in MVP-based test rooms - to obtain continuous feedback. Internal “champions” also help to transfer enthusiasm to the rest of the team.

On the other hand, a genuine transformation requires sufficient resources and a willingness to invest: without a dedicated contact person, a budget for licenses and the necessary data protection and data security measures, initial experiments with free tools will fail to deliver the major added value. In addition, the entire effort behind a technical solution is often underestimated - not just the implementation, but also the necessary “digital homework” such as the precise analysis of your own processes, responsibilities, audit rules and contractual red flags.

Correctly addressing the problem is also key: instead of blindly purchasing technologies, you need to address specific pain points that are relevant to several employees. Only on this basis - after standardizing templates and optimizing processes, as shown in the example of an influencer contract process - can the subsequent technical solution really take effect. Finally, data protection and data security hurdles should not be neglected and require additional investment and care to ensure that the AI solution works in day-to-day business.

Darko: If you take a look now, David, what is the common denominator among these companies that are already using AI for contract review?

David: For all companies that already use AI for contract review, three core factors are common denominators throughout the entire process:

Darko: People instinctively ask themselves whether traditional lawyers are still needed at all? And if we put a positive spin on the question, Alisha and David, what mindset or skills need to be implemented now to avoid being left behind?

Alisha: Even if individual activities may already be called into question today, there will certainly still be a large number of lawyers in the future. Other skills than pure legal knowledge are now crucial for their continued existence: At the interface between disciplines, they must develop interdisciplinary skills, i.e. have a certain basic understanding of technical issues and at the same time recognize where their own legal boundaries lie and where they need to approach experts from other disciplines. It is precisely this mutual recognition and respect for the skills of other disciplines that is essential. As AI can take over many routine tasks, the more complex and strategic thinking of lawyers becomes all the more relevant and exciting - discovering and bringing together potential far beyond the schema F checking of contracts remains a human domain.

An iteration mindset is just as important: lawyers should have the confidence to share imperfect results, obtain feedback and iterate step by step towards a better solution instead of expecting to deliver perfect work from the outset. This is accompanied by critical thinking in order to be able to constantly question AI results - in other words, to keep checking: “Is that really what I wanted to say?”

And last but not least, they have to get used to using AI on a daily basis: simply trying it out, learning where AI helps and where its limits lie in order to identify personal use cases on the one hand and improve prompting on the other. This constant use will train the new “AI muscle” and make lawyers indispensable in the future.

David: Even though Alisha has already summarized the topic very nicely, three central mindsets and skills can be identified that lawyers need now in order not to be left behind:

Darko: I have one more question: it's purely hypothetical, but could also become realistic very quickly. Now Alisha comes up to you, David, and says: “Hey, I have a customer here with a very specific use case and we'd like to tackle it with you from Legartis.” How long would it take to actually bring this use case live?

David: A proof of concept can be implemented in just a few weeks if it is clearly defined in advance what exactly is to be done, what challenge is to be solved and what goal is to be achieved. However, you should allow at least six months for the complete end-to-end implementation, where all stakeholders and users are onboarded and the solution is fully functional.

You feel ready for a change and would like to have your use cases reviewed for the contract review at Legartis? Then take a look at our special Legartis license offer and book a meeting.

Questions and Answers

Other tools produce very good results when used in compliance with data protection regulations. What justification do you recommend for implementing special tools? The GDPR topic or are there other advantages, especially for very small legal departments?

Alisha: Although many use cases can already be covered with broad-based AI tools like ChatGPT, Gemini, or Microsoft offerings, we still recommend using specialized tools. In our experience, these expert systems incorporate much more concrete functions specifically tailored to legal work, and often include legal expertise – of course, to varying degrees depending on the tool chosen. This added depth goes beyond the capabilities of general, mass-capable platforms and corresponds to the usual pattern in the software landscape: on the one hand, there are broadly positioned solutions that serve many scenarios, and on the other hand, expert systems that make significantly more progress in their field. For particularly complex tasks such as a comprehensive contract review, general offers quickly reach their limits; specialized tools are the only ones that can provide the desired level of precision and depth.

David: Although widely available AI tools like Gemini or ChatGPT are frequently used and can even perform simple contract reviews, the use of specialized solutions is recommended for two main reasons: first, quality assurance for recurring review tasks. While you can have ChatGPT check a contract for 20 points once, there is no guarantee that the result will meet the same standard after a thousandth run. Specialized tools like Legartis link each individual test point to a specific prompt and a corresponding test set. This ensures the consistency and accuracy of the results across all test cycles – a decisive advantage, especially for small legal departments with frequently recurring use cases.

When AI and Legal Design are combined, how can legal departments be enabled to exert strategic influence on companies? Also, ein Business Enabler innerhalb eines Unternehmens zu werden. How do AI and Legal Design work together here?

Alisha: AI serves as a technical enabler: It allows us to "do many more things at the same time, do things with a higher quality," provided we "use it correctly" and place it exactly "where we want it." Legal Design, in turn, forms the strategic path to leveraging this potential in a user-oriented way: With a "user-oriented problem approach" and the Legal Design "building kit," we develop the necessary strategy and set up technology and people in such a way that the desired added value is created.

Legal departments often "unconsciously sit on a huge treasure" – the data from contracts and inquiries. As a "hub," they process information that, when "properly used" and initially made transparent and prepared for the target group, can provide enormous strategic value for the entire company. To control, we need technology – AI.

In combination, AI acts as an enabler, while Legal Design creates the framework as a toolbox: Together, they are "perfectly suited to deliver strategic value for the entire company."

Darko: Thank you Alisha for your insights. It is indeed always valuable to have such expertise in a webinar. Many thanks David. For conveying your 8-year AI development expertise. It is a lot of heart and sweat that you invest in the areas of AI and Legal AI. Thanks to the participants. Thank you.