LEO as a legal AI assistant at a law firm

In the legal profession, time is money, and that time is often spent sifting through information. Lawyers must "know the law," but the reality is an overwhelming amount of laws, rules, and court decisions that constantly change. Manually finding the right case law or reviewing stacks of contracts in an acquisition file can take days or weeks. This routine work is not only time-consuming but also prone to errors – people can overlook important details after reading dozens of documents. Moreover, clients are becoming more demanding: they expect quick answers and want to pay less for hours of research that require little creativity. For the law firm, it was a challenge to remain efficient: senior lawyers were stuck in endless reading and searching tasks, while their added value lies in strategic advice and negotiation. Training junior lawyers is also intensive, as they need a lot of time to learn how to use all relevant sources due to the learning curve. In short, there was a need for a tool to conduct legal research faster and automate repetitive tasks without compromising accuracy.

The firm implemented LEO, a custom-trained legal AI agent. LEO was connected to a legal knowledge base of legal texts, regulations and case law, plus relevant parliamentary documents to understand legislative history and context. Additionally, LEO can analyze documents provided – for example, a set of contracts or a case file provided by the client. The workflow is as follows: a lawyer starts a chat with LEO via a secure portal and creates a client file. He/she can, for example, ask "Which case law is relevant for summary dismissal in <case>?" or "Check this agreement for unusual risks for our client." LEO understands the question, quickly searches its legal corpus, and provides an answer in clear Dutch, often with a list of findings (articles and case numbers) included. For more complex questions, LEO generates a structured advice with points of attention. Thanks to the integration with the file and document management system, LEO can also draft a first version of legal documents directly. For example: filling in a draft summons based on templates and specific case information, or developing an advice letter to the client with the findings. The AI also takes over administrative tasks: linking documents to the correct file, scheduling deadlines (e.g., if a certain document is still missing or action is required). During an M&A due diligence, the team uses LEO to analyze all submitted contracts and financial documents for red flags: the AI scans hundreds of documents in minutes and presents an overview of potential risks, from change-of-control clauses to unusual obligations. That saves the team a great deal of manual reading and searching, while the lawyer decides what happens with the findings.

More on what LEO does for lawyers and in-house counsel: LEO, the legal AI assistant.

Added value of AI (LEO): LEO elevates legal services to a higher level in multiple areas. Firstly, speed: Answers and first document versions are now almost immediately available. Case law research that used to involve a lot of digging now produces a first overview quickly. This means faster turnaround times for clients and room to handle more cases. Secondly, thoroughness: LEO searches the sources systematically and cites its source with every answer, so the lawyer can check what was taken into account. This increases the quality and completeness of advice. Thirdly, efficiency and cost savings: By automating repetitive tasks, expensive hours of experienced lawyers can be spent differently. At this firm, that means less time spent reading and typing and more time advising, which reduces workload and leaves room to quote more competitively. Fourthly, consistency: All documents are prepared in a uniform format and language. Standard clauses in contracts and letters follow the best (LEO-suggested) template. This ensures consistent quality regardless of which employee works on it. Junior lawyers also climb the learning curve faster through LEO: they immediately see what an experienced letter should look like and why certain case law is relevant, as the AI explains this. LEO thus acts as a kind of mentor in addition to being a tool.

Since the introduction of LEO, the productivity of the legal team has increased enormously. Research work that used to cost many billable hours now goes faster, which translates into a lower bill for the client. This makes the firm more competitive and gives lawyers the opportunity to take on more cases in the same time. In practice the firm sees time gains on document drafting and research work. Clients notice it in the speed: advice that used to come after a week is now often on the table within a day. Moreover, the quality has remained high or even improved – there are rarely complaints or missed points, as LEO's thorough approach reduces human errors and overlooked points. For example, in a due diligence for an acquisition the firm used LEO to work through the contracts supplied and put together a risk overview for the client. The partners note that their lawyers now have more time for actual legal thinking: strategy, argumentation, client contact. Routine tasks are confidently left to LEO. The work-life balance at the firm also improved somewhat – working overtime to complete research is less often necessary. Finally, the firm positions itself through this innovation as a leader in legal tech, attracting new talent and impressing clients. In practice, LEO proves that AI can do much of the time-consuming legal work faster and better than was previously possible, while the human lawyer continues to fulfill the crucial creative and empathetic tasks. This allows the firm to offer its clients faster, cheaper, and yet extremely careful legal advice – a compelling proposition in a traditional industry.