Building legal AI that lawyers can trust with Alexander Kardos Nyheim, Thomson Reuters

Law and legal services are among the toughest tests for artificial intelligence. In this episode of The Influencers, Alexander Kardos Nyheim joins Leo von Gerlach to discuss why legal work demands AI models that are pre-trained differently on top of their general‑purpose abilities.   Drawing on his experience founding Safe Sign Technologies and now leading machine learning research at Thomson Reuters, Alexander explains why general purpose AI’s emphasis on fluency is no proper match for the finesse of legal drafting. He explores the importance of domain‑specific training, authoritative legal data, and models grounded in legal reasoning rather than surface‑level coherence. From the role of foundational models to the future of legal‑specific AI, this conversation offers practical insight for anyone navigating the adoption of AI in legal practice.   About Thomson Reuters:  Thomson Reuters is an AI and technology company empowering professionals with trusted content and workflow automation to know now and navigate what’s next   About Hogan Lovells: Our Digital Transformation offering helps clients navigate the full breadth and depth of the seismic forces impacting their organizations – and today’s society. Leveraging team members across our 2,800 lawyers on six continents, we provide integrated advice on the multifaceted regulatory, governance, and enforcement issues and risks that arise across the many technologies and trends that comprise the new digital future.   If you enjoyed today's episode, we'd love for you to rate and review us. Your feedback helps us reach more listeners and continue bringing insightful conversations on digital transformation and law. Thank you for your support!