European law firms have a new ally: JUPUS, a German startup that just banked €13 million to drive its artificial intelligence into professional practices. The Series A round was led by Semapa Next, with NRW.BANK joining through its NRW.Venture fund and existing investors Acton Capital and High-Tech Gründerfonds participating.

An AI secretary for lawyers

Billed as “Europe’s first AI secretarial service built specifically for law firms,” JUPUS’s platform automates a chain of tasks once the exclusive domain of human assistants: answering client calls, structuring inquiries, preparing case files and drafting legal documents. The company estimates savings of more than 70 hours per month for a law firm.

The numbers signal rapid adoption: over 2,000 lawyers already use the system and more than 2,000 new cases are processed daily by the AI. In 2025, annual recurring revenue (ARR) quadrupled, headcount tripled and the user base—along with the volume of cases handled—more than doubled.

The law firm paradox

The backdrop explains the traction. In Germany, over the past three decades the number of newly trained legal assistants has plunged by over 70%, while practicing lawyers have tripled and administrative workloads have ballooned. Small and mid-sized firms—which dominate the European market—have been hardest hit by the shortage of support staff. JUPUS aims to fill that gap with a vertical AI capable of handling routine tasks autonomously, freeing time for client advisory and actual legal work.

AI and legal practice: the privacy battle

AI’s expansion into the legal sector brings tough questions about confidentiality and data sovereignty. Co-founder and lawyer René Fergen notes that generic models offer support tools, but legal professionals need solutions tailored to specific regulatory, privacy and professional requirements. JUPUS has built its offering around those needs, yet the deployment model remains a sticking point: a cloud-based service—however specialized—raises questions about where and how sensitive data covered by attorney-client privilege travels.

This is where the evaluation becomes relevant for law firms or IT departments weighing the trade-off between speed of adoption and infrastructure control. AI-RADAR provides analytical frameworks to compare on-premise and cloud deployments of Large Language Models, focusing on concrete parameters such as TCO, inference latency and GDPR compliance. In this case, JUPUS’s managed-service approach accelerates penetration but may face resistance in settings with strict data-localization requirements.

Toward a more automated Europe

With the fresh capital, the startup will invest in evolving its AI capabilities and deepening its presence among small and mid-sized firms across the continent. The market response and investor interest confirm that legal automation is moving beyond the experimental phase. The open question is whether an “as a service” model will prevail, or whether self-hosted solutions will emerge that can guarantee maximum control without sacrificing automation quality.