Summary
Ryan Anderson, co-founder and CEO of Filevine, joins Ceschino to discuss how AI is moving from a one-question-at-a-time tool to an always-on, matter-aware system for legal work. The conversation covers a broad range of topics including the advent of agentic AI, how AI can anticipate legal teams’ needs, simulation of judicial reasoning for litigation strategy, and why anti-hallucination tooling is becoming not just essential, but possible in legal practice.
Takeaways
AI in law is shifting from occasional use to an always-on layer that anticipates needs before lawyers ask.
An agent is an AI system tasked with pursuing a goal through multiple steps, not just answering a single prompt.
Multiplayer, agentic systems could coordinate work across an entire matter or case team.
AI can uncover insights teams may not have found manually, rather than simply accelerating existing tasks.
Research agents can simulate how different judge archetypes might reason through a case.
AI tools such as LOIS for Word can plan, ingest case files, analyze prior drafts, and help produce stronger briefs.
Anti-hallucination technology is becoming essential for verifying legal research and citations.
Legal AI must detect fabricated cases, misquoted cases, and quotes taken from the wrong part of an opinion.
Lawyers still need professional judgment even as AI improves search, drafting, and citation checking.
Junior lawyers may struggle to develop foundational research and writing skills if they rely on AI too early — the profession needs a plan for how to train new entrants.
The future of legal AI may involve many highly specialized models instead of one dominant system.
AI is more likely to expand lawyer capacity than replace lawyers entirely.
Chapters
00:00 AI in law and the “cold fusion” moment
05:30 Transforming legal work, not just accelerating it
08:45 What is an AI agent?
13:18 Developing anti-hallucination technology in legal AI
16:29 Three common types of legal AI hallucinations
21:15 Drawing inspiration from the people who shape law and justice
23:57 Testing anti-hallucination tech and legal aI
26:26 The importance of context engineering
30:19 Ryan’s controversial opinions on the legal tech industry
30:42 Controversial opinion #1: millions of custom AI models
31:24 Controversial opinion #2: AI won’t replace lawyers
32:36 Controversial opinion #3: AI creates a training challenge for the legal profession
39:06 Ryan’s fun fact and closing thoughts
Keywords
AI, legal tech, legal AI, AI hallucinations, legal research, AI agents, agentic AI, legal innovation, justice, law practice, AI models, legal ethics