How AI is Transforming Legal Discovery — A Conversation with Aaron Patton
Down in the Bunker | Season 2, Episode 2
Legal tech is moving fast, and nowhere is that more apparent than in the world of eDiscovery. In this episode of Down in the Bunker, we sat down with Arron Patton, lawyer, solutions architect at Everlaw, and one of the most thoughtful voices at the intersection of law and technology — to explore what artificial intelligence is actually doing to legal discovery, and what it means for the profession’s future.
Who Is Aaron Patton?
Aaron isn’t your typical tech evangelist. He trained as a lawyer, cut his teeth in database administration at Netscape during the dot-com era, and has spent years working in discovery consulting. That rare combination, legal rigour meeting deep technical fluency, makes him exceptionally well-placed to speak honestly about where AI is delivering real value, and where the hype still needs to catch up with reality.
Today he works as a solutions architect at Everlaw, helping law firms and legal teams navigate sophisticated software and AI-powered tools. He’s seen what works, what doesn’t, and what questions people aren’t asking but should be.
From TAR to AI: A Shift in Scale
For years, technology-assisted review (TAR) has been the benchmark for making sense of the enormous volumes of documents that surface in complex litigation. It was, and remains, genuinely transformative. But as Arron explains, AI is now enabling something qualitatively different: the ability to interrogate large datasets through natural language, rather than relying solely on keyword searches or trained review sets.
Tools like Everlaw’s AI assistant allow a legal professional to ask a plain-English question and receive a contextualised answer drawn from the document corpus — not just a list of hits, but a response that draws meaning from the material. For anyone who has spent hours crafting Boolean search strings, that shift feels rather significant.
The Validation Problem
Arron is careful not to oversell the moment, though. One of the most important points he makes is about validation. AI outputs are only as trustworthy as the processes used to check them, and legal work, where accuracy can carry enormous consequences, demands a particularly high standard of scrutiny.
This creates an interesting dynamic. As AI handles more of the initial review work, a new kind of role is emerging: professionals focused specifically on evaluating and validating AI-generated outputs. It’s less about reading every document and more about understanding whether the tool got it right, why it reached the conclusions it did, and where its blind spots might lie. That requires a blend of legal judgement and technical literacy that doesn’t yet have a neat job title, but is increasingly in demand.
Context Is Everything
Perhaps the most practically useful thing Aaron says is this: context matters enormously when working with AI. A tool might surface a highly relevant document or offer a confident-sounding summary, but without understanding the surrounding circumstances — the legal theory, the jurisdiction, the specific dispute, a practitioner can easily misread what they’re looking at.
This isn’t a criticism of AI so much as a reminder that these tools amplify the judgement of the person using them. A skilled, context-aware lawyer with a good AI tool will outperform both the tool alone and the lawyer alone. The technology raises the ceiling; it doesn’t replace the expertise underneath it.
What This Means for Legal Careers
There’s understandable anxiety in the profession about what AI means for jobs. Aaron’s take is measured. Some tasks that currently occupy significant billable time, first-pass document review, for instance, will likely be handled increasingly by AI. But that doesn’t mean the lawyers doing those tasks simply disappear; it means their time gets redirected.
The question is whether the profession adapts quickly enough. Firms and practitioners who invest in understanding how these tools work, not just how to switch them on, but how to evaluate their outputs critically, will be far better positioned than those who either resist the change or accept AI outputs uncritically.
The Bigger Picture
What makes this conversation particularly valuable is Aaron’s refusal to land in either the utopian or the dystopian camp. AI in legal discovery is genuinely useful, genuinely improving, and genuinely risky if misunderstood. The firms that will navigate this transition well are those that treat AI as a powerful collaborator requiring careful oversight, not a black box to be trusted blindly, and not a threat to be dismissed.
For anyone working in legal tech, practising law, or simply curious about how AI is reshaping professional services, this episode is well worth your time.
Spotify:
#LegalTech #AI #eDiscovery #Everlaw #DownInTheBunker #Podcast #ArtificialIntelligence #LawAndTech #Forted #TheFortedBunker