LegalAI

AI Privacy for Lawyers: Your Client Data Has Entered the Chat

Lawyerist
Lawyerist•May 21, 2026

Why It Matters

If lawyers ignore AI‑driven privacy risks and resist scientific updates, client confidentiality and wrongful‑conviction safeguards will deteriorate, undermining the legitimacy of the legal system.

Key Takeaways

  • •AI threatens lawyer-client confidentiality; privacy expectations must evolve.
  • •Conviction review units expose wrongful convictions using new evidence.
  • •Loyalty can blind firms to necessary “pruning” of relationships.
  • •Courts lag behind scientific advances, hindering post‑conviction relief.
  • •Collaborative models with innocence projects amplify exoneration outcomes.

Summary

The episode tackles two intersecting concerns for modern law practice: the erosion of client‑data privacy in an AI‑driven world and the growing imperative to revisit past convictions. Host Bernardet and guest Sunny Eaton use a recent Women in AI conference discussion as a springboard to argue that lawyers must redefine privacy expectations before AI tools inadvertently expose privileged information. Key insights include a metaphor about "pruning" loyal but outdated relationships—whether with clients, vendors, or employees—and how that mindset applies to legal workflows. Eaton explains the function of Nashville’s Conviction Review Unit, noting that only about 85 such units exist nationwide, yet they have already secured six exonerations and have three cases pending, underscoring the tangible impact of systematic case re‑examination. Memorable moments come from Leticia’s analogy that loyalty offers "a hundred reasons to keep feeding it," and Eaton’s description of new DNA technologies that can overturn decades‑old forensic conclusions. The conversation also highlights the friction between rapidly evolving scientific methods and courts that remain notoriously slow to adapt. The broader implication is clear: law firms and prosecutorial offices must adopt proactive privacy safeguards, embrace technological advances, and cultivate internal channels for ethical self‑review. Failure to do so risks both client confidentiality breaches and the perpetuation of wrongful convictions, eroding public trust in the justice system.

Original Description

Wrongful convictions and AI privacy may seem like separate issues, but both raise the same uncomfortable question: what happens when the legal system relies on old assumptions in a changing world? In episode 618 of the Lawyerist Podcast, Zack Glaser talks with Sunny Eaton about conviction review work, evolving science, and why lawyers should not be so quick to surrender privacy expectations in the age of AI. 
Sunny shares how her work in the Nashville District Attorney’s Office focuses on reviewing old convictions, identifying new evidence, and helping correct cases where the system may have gotten it wrong. She explains why changing science, including advances in DNA, trauma research, and bias studies, can matter deeply when reviewing criminal convictions. 
The conversation then turns to AI, attorney client privilege, client data, and the growing role of data brokers. Zack and Sunny explore whether information shared with AI tools should automatically lose privacy protection, and why lawyers may need to make stronger arguments before courts accept that assumption. 
If you are concerned about AI in law firms, client confidentiality, or the future of privacy rights, this episode challenges lawyers to think harder about what should remain private and why it is our job to make the government work for it.  
Have thoughts about today’s episode? Join the conversation:
_______________________________________________________________________
Subscribe to the Lawyerist Podcast and explore more episodes:
Learn more about Lawyerist’s Coaching Program:
Read the Small Firm Roadmap Revisited:
Visit our Web Store for more resources:
Chapters / Timestamps: 
00:00 Introduction 
01:25 Pruning Your Practice 
04:15 When Loyalty Gets in the Way 
06:25 Meet Sunny Eaton 
08:45 Conviction Review Units 
09:50 Reviewing Old Cases 
11:25 Building Trust 
12:40 What Cases Qualify 
14:35 Changing Science 
16:00 DNA, Trauma, and Bias 
18:05 Case by Case Review 
19:25 Bad Science and Old Convictions 
20:40 AI and Privacy 
22:00 Lawyers, Therapists, and Privacy 
23:20 Giving Up Too Quickly on AI Privacy 
25:30 Reasonable Expectations of Privacy 
26:15 Client Data and AI Risk 
29:15 What Email Teaches Us 
31:00 Attorney Client Privilege and AI 
33:20 Personal vs. Workplace AI Use 
34:00 Old Rules, New Technology 
35:45 Make the Government Work for It

Comments

Want to join the conversation?

Loading comments...