Tailored AI assistants for the work of a specialized law firm

BKL Fischer Kühne + Partner advises on wealth succession, banking and capital markets law, and tax and foundation law, among other areas. Eight lawyers work at the offices in Bonn, Hamburg and München. For Dr. Bastian Schmidt-Vollmer, partner at the Hamburg office, the question was not whether AI would be used at the firm. What mattered was the form in which it can support specialized work.

Ready-made legal AI products did not convince BKL for this use. The firm wanted a tool that can be adapted to its practice areas, client structure and quality standards, rather than forcing its work into a predefined standard process.

At the same time, the secure handling of client information was a clear prerequisite for the selection. BKL attached importance to hosting in Germany, Zero Data Retention for model requests, no use of customer data for model training and a solution that meets the requirements of § 203 StGB. For the firm, these points were not an added benefit but the foundation for using AI in client work at all.

BKL has created its own AI assistants in Logicc. Each has a defined area of use and a system prompt that reflects the firm's practice area, way of working and quality standards.

The difference from a general AI chat becomes tangible when drafting a legal brief, for example: “Dr. T.” is geared to this task and to the firm's professional context. He supports with structure and the first draft; the legal review and the decision remain with the lawyers. BKL can further develop the assistant's instructions itself when practice shows what needs to work better.

Beyond the assistants, the firm has implemented its own integration for matching tax assessments in the context of closed-end funds. With roughly 2,000 investors, this use case extends the use of Logicc beyond work in the chat.

Today, BKL uses Logicc for first drafts, research and the structured preparation of complex matters, among other things. The firm's own AI assistants support different professional tasks. The assessment matching adds another use case, implemented specifically for the firm.

The rollout also shows that access alone does not mean everyone on the team uses AI equally. BKL therefore treats the setup not as a completed rollout but as a tool that the firm keeps adapting and embedding in everyday work.