The Intersection of AI, Legal Reasoning, and Access to Justice with Damien Riehl and Jacqueline Schafer

Chapter 1: Introductions

Jack Newton: Jackie, Damien, thanks for joining us this morning. I’m really looking forward to this conversation. We’re gathered here in Chicago in the Wicker Park neighborhood—super cool to be here with both of you. I’d love to start with brief introductions of what you do at your companies.

Jacqueline Schafer: Absolutely. I’m the founder and CEO of Clearbrief. I’m a career litigator. I started in big law in New York, spent years in government doing appellate work and complex litigation, and was also in-house counsel. Clearbrief is a Microsoft Word add-in that interacts with your documents to keep any evidence you’re referencing visible. You can create timelines and summaries, check everything before filing, and share it with the court. We’re used by the largest global law firms down to solo practitioners, courts, and government agencies.

Damien Riehl: I’m Damien Riehl with vLex. My title is Solutions Champion, which is a made-up title for a made-up job, but it might be the best job in legaltech. I build in the morning with my product team and talk to customers in the afternoon. There are a billion legal documents—cases, statutes, and pleadings. We use Large Language Models (LLMs) to extract data and perform generative tasks like drafting complaints and answers. We refine that “public oil” with LLMs.

Jack Newton: Chicago is such a hub of legal tech innovation—home of the ABA and the Tech Show. What makes it special for you?

Jacqueline Schafer: I’m a jazz musician, so I associate Chicago with historic jazz clubs.

Damien Riehl: I’m a Midwesterner through and through. Chicago is just a one-hour flight from my house in Minnesota. It’s got a great vibe—rough and tumble like New York, but Midwestern polite.

Chapter 2: Real applied AI vs. The hype phase

Jack Newton: We’ve moved past the hype phase. You are both in the trenches of real applied AI. Jackie, what are you seeing as high-impact applications and current limitations?

Jacqueline Schafer: We recently redesigned our UI because customers were frustrated with open-ended prompting. In litigation writing, you start with factual documents that you must search and organize. Checking everything over has become a huge focus due to news about hallucinations and sanctions. There was a recent case where an entire legal team was sanctioned because an associate used an AI tool that hallucinated citations that weren’t vetted. We focus on “grounding” the user, showing them exactly how to check their work with hyperlinked sources directly inside Word.

Damien Riehl: Regarding the Gartner Hype Cycle, I think it’s an individual journey. While some are on the “slope of enlightenment,” many solo lawyers haven’t even reached the “peak”. At vLex, we reduce hallucinations by connecting “symbolic AI”—good old-fashioned, deterministic AI—with LLMs. This is Retrieval-Augmented Generation (RAG). If you have a billion documents and use symbolic AI to tag them, the hallucination rate is close to zero because you’re retrieving ground truth cases before generating text.

Chapter 3: Practical applications for the daily workflow

Jack Newton: If you could recommend three or four things a lawyer should leverage AI for today to stay competitive, what would they be?

Damien Riehl: Don’t just think about “generating” text; think about “extracting” it. Use it to pull facts and summarize documents at superhuman speed. It can provide a better first draft than a junior associate.

Jacqueline Schafer: There are three core things. First, automated timelines. You can dump in documents and instantly get a hyperlinked timeline in Word with citations. Second, Table of Authorities. Why do this manually? Classic AI can identify every case you cited and build the table for you. Third, hyperlinking. Providing a judge with a brief where they can click a link and immediately see the supporting case or evidence is incredibly powerful.

Damien Riehl: It shifts from “trust but verify” to “trust and verify”. When you see the ground truth text 100% of the time because it’s a database search rather than an LLM generation, you build a virtuous cycle of trust.

Chapter 4: The future of AI agents

Jack Newton: Where are we heading in the next five years?

Jacqueline Schafer: We’re moving toward “agents”. The challenge is ensuring lawyers feel comfortable with what an agent can access and execute. Reputation is everything in law; a single hallucinated brief can impact a career.

Damien Riehl: There are two types of agents: LLM agents, which are autonomous and a bit magical, and symbolic agents, which are reliable expert systems like TurboTax. At vLex, we use symbolic agents to do the repetitive work—identifying claims, facts, and parties—deterministically. Then we sprinkle LLM “pixie dust” on top for custom research questions. By 2027, we might see 50 to 75 years of progress because LLMs are learning how to code themselves.

Jacqueline Schafer: The hardest part isn’t the tech; it’s bringing the people along. I still hear a lot of resistance at conferences. We have to make this human-centered and show people the “why”.

Chapter 5: The death of the billable hour?

Damien Riehl: The risk for lawyers who don’t adopt this is that the phone simply stops ringing. Clients, especially in-house counsel, are starting to use these tools themselves for initial analysis. They won’t tell you they’re not hiring you because you aren’t using AI; they’ll just stop calling.

Jacqueline Schafer: It’s an opportunity for lawyers to become better businesspeople. You have to anticipate client needs and market yourself as future-ready.

Damien Riehl: There are two future worlds. In World One, the associate-to-partner ratio shrinks because you don’t need humans for 200 hours of manual work. In World Two—the better world—lawyers use AI to shrink unit costs and tap into the 92% of legal needs that are currently unmet because we are too expensive. This “latent market” is a trillion-dollar opportunity.

Jack Newton: Leveraging AI gives lawyers the freedom to walk away from the billable hour and offer predictable pricing.

Damien Riehl: “The slowest lawyer wins the race” only applies to hourly billing. For contingency, flat fee, or subscription models, efficiency increases profit margins.

Chapter 6: Resources and recommendations

Jack Newton: What resources do you recommend for listeners who want to dive into AI?

Jacqueline Schafer: Follow Ethan Mollick on LinkedIn. He’s a Wharton professor who posts every time a new model drops. Also, deputize someone at your firm to experiment with technology and learn about security.

Damien Riehl: I second Ethan Mollick; his book Co-Intelligence is fantastic. But ultimately, asking how to use AI is like asking what book to read to learn how to swim—you just have to swim. Put a post-it on your monitor that says “Can AI help with this?” and then go try the tools.

Jack Newton: Damien, Jackie, thanks so much for joining me today. I really enjoyed this conversation.