What AI Is Already Doing to the Legal Industry

Episode Summary

Executive Summary: The episode examines how AI is reshaping legal work, especially big law and litigation. Guest Joel Wertheimer argues AI will automate grunt work like research, document review, and formatting, but increase the value of client management, oral advocacy, and case sourcing. The conversation weighs effects on billable hours, partner leverage, junior training, access to justice, and the possibility that legal services become cheaper, more abundant, and more litigated.

Main Topics: AI as a productivity tool for lawyers (Priority: 5/5): The discussion centers on how newer models, especially OpenAI’s O3/O3 Pro, can accelerate legal research, find relevant cases, summarize records, and reduce time spent on repetitive tasks like discovery responses and formatting. Big law economics and leverage (Priority: 5/5): Joel explains the associate-to-partner leverage model, Cravath-scale pay, billable-hour incentives, and how AI could alter who captures value inside firms by reducing junior labor needs and boosting partner productivity. Training, junior lawyers, and the future of legal work (Priority: 4/5): The guests discuss how law school does not teach the practical mechanics of litigation, why junior associates learn on the job, and how AI could either weaken that apprenticeship or make junior work more efficient. Billable hours vs. alternative fee models (Priority: 4/5): The episode considers whether AI will erode the traditional hourly billing model and whether flat-fee or other alternate fee arrangements will become more common as firms seek to preserve margins. Access to justice and Jevons paradox in law (Priority: 5/5): AI may lower the cost of taking cases, enabling more contingency work, more pro se litigation support, and more representation for lower-value or underrepresented claims, potentially increasing overall legal volume. Client privacy, risk, and adoption barriers (Priority: 3/5): The conversation notes confidentiality concerns around uploading sensitive data like medical records or client documents to AI systems, especially in large firms handling high-stakes corporate matters. Future of legal services, case acquisition, and advertising (Priority: 3/5): The speakers speculate that AI will make sourcing clients and cases more valuable, reshape SEO and advertising, and may intensify competition among plaintiff firms and legal marketers.

Key Arguments: AI is already useful for legal research and can identify relevant cases and issue spots faster than older tools, especially when compared with earlier GPT models. Routine legal work—discovery review, privilege screening, document formatting, and template drafting—is the most obvious target for automation. Big law’s leverage model depends on junior associate labor; if AI reduces that labor, firms may shift more value to partners or use fewer associates. Lawyers are still needed for judgment, client counseling, deposition strategy, oral argument, and validation of AI-generated research. AI could reduce the cost of bringing cases, making it easier for plaintiffs’ firms and legal aid lawyers to take more matters and serve more clients. The legal market may expand rather than shrink if lower costs create more demand for legal services, more lawsuits, and more self-help litigation. Confidentiality and data security remain major constraints on AI adoption, particularly when handling medical records or sensitive corporate information. Client-side incentives may favor AI if it reduces costs, but high-stakes clients will remain cautious about model quality and data handling. Case acquisition and marketing may become even more important as AI improves efficiency and increases the profitability of high-volume legal practices.

Data Points: Stock Movers length: five minutes or less - Described in the Bloomberg promo at the start of the episode Bloomberg journalist/analyst network: 3,000 journalists and analysts - Mentioned in Bloomberg’s podcast promotions Discovery billing increment: 6-minute increments - Joel describes how law firms track associate time Associate bill rate: $700 an hour - Example of a big law associate’s billing rate Annual associate hours: 2,000 hours a year - Used to estimate revenue generated by an associate Associate revenue generation: $1.4 million - Joel’s rough calculation of annual revenue per associate at $700/hour and 2,000 hours Associate compensation with taxes: $500,000 - Approximate total cost to firm after compensation and taxes Partner profit per extra associate: $900,000 - Joel’s estimate of profits attributable to leverage Illustrative partner profits: $3.6 million - Example using a four-to-one partner-to-associate leverage ratio Partner billing rate example: $2,000 an hour or $1,500 an hour - Illustrative partner bill rates in big law Associate-to-partner effective value: $5,000 an hour for partner and $300-$400 for associate - Joel’s point about how value may really accrue behind the scenes Potential pro se share of cases: 20-25% - Estimate Joel cites for current cases handled by self-represented litigants Contingency fee in New York: one third - Joel describes standard plaintiff-side contingency arrangements Short-form podcast duration: five minutes or less - Reinforced in Bloomberg promo segments Guest’s timeline for starting firm: January 2021 - Joel says he started his own firm then Example legal research model: O3 / O3 Pro - Referenced repeatedly as especially strong for legal research Model evaluation example: one Georgetown professor running exam questions - Used to illustrate AI performance improvements over time

Pivotal Quotes: "I think it's going to be hard for junior lawyers at big law firms to get their sort of that training." — Joel Wertheimer: On how AI may affect on-the-job learning and apprenticeship in firms "The lawyers who can make them feel safe and let them understand what's happening to them and what their odds are, those sorts of things... that's going to go up and up." — Joel Wertheimer: On which legal skills become more valuable as AI automates routine work "I think it's going to be fewer and what they do will change a lot." — Joel Wertheimer: His answer to whether there will be more or fewer lawyers in ten years

Implications: AI is likely to compress routine legal work, raise output per lawyer, and shift value toward senior judgment, client trust, and case sourcing. Expect more legal volume, stronger tools for smaller firms and litigants, and growing pressure on billable-hour economics.