AI Tools for Practicing Lawyers

The episode delivers substantial, non-obvious insights about evidence law and AI deepfakes. Judge Grimm explains specific evidentiary challenges (the 'liar's dividend,' authentication thresholds, prejudice tests) and proposes concrete framework solutions (Rule 901C). However, there is meaningful filler including extended procedural explanations of the Rules Enabling Act process and some repetitive throat-clearing about AI capabilities.

“What's happened is we've democratized fraud. We've made it available at an enormous scale.”

“The liar's dividend. The person who lies and says, oh, that's a fake, everybody knows these deep fakes are out there, you know, we you can't rely upon that, then you get a jury that doesn't know what they can rely on.”

Judge Grimm's framing of deepfakes as an evidence problem rather than solely a technology problem is thoughtful, and the 'liar's dividend' concept provides a genuinely fresh lens on jury fact-finding. The proposed Rule 901C represents original thinking about burden-allocation. However, the core insight - that existing evidence rules are technology-agnostic and adaptable - is not particularly novel, and much of the discussion applies familiar legal doctrine to new facts.

“Has AI fundamentally changed the law of evidence, or has it just made it easier for us to revisit principles that have always been there?”

“The rules of evidence in the United States and probably elsewhere, they're really technology agnostic.”

Judge Paul Grimm is exceptionally well-calibrated for this topic: 25+ years on the federal bench, recognized expert in evidence and ESI, published extensively on law-technology intersections, and currently directing judicial education at Duke Law. He brings practitioner-level credibility combined with scholarly depth. This is not a career podcast guest but a genuine authority who has shaped judicial thinking on the subject matter.

“Judge Grimm served for more than 25 years on the federal bench, including as the United States District Judge for the District of Maryland, where he became widely recognized for his work on evidence, civil procedure, and electronically stored information.”

“I've been fascinated by the challenges that digital information presents from an evidentry point of view uh for quite some time. I happened to have the good fortune to join the federal court back in 1997, right when digital information explosion had occurred.”

The episode provides concrete examples: State v. Pulka (Washington 2024), the Maryland facial recognition statute, State v. Albisu (Florida 2025 VR scenario), a California Superior Court sanction case involving fabricated images, and hypotheticals about voicemail deepfakes. Judge Grimm cites actual proposed rules (901C, Rule 707) and real companies (Learned Hand). However, many examples are fact patterns rather than detailed metrics or dollar figures, and some case citations lack full development.

“In the state, in the case of state of Washington versus Pulka, uh, which was decided in 2024 as a state court case in uh King County Superior Court in Washington State, it was a criminal case.”

“There's another one, and I'm gonna say something too. There's an existing federal rule right now, and this is flying under the radar scope. This is a fairly new rule. It's rule of evidence 107, just was adopted in the last couple of years, and it's called illustrative evidence.”

The hosts (Ron Drescher and Heather Gardner) ask competent setup questions and allow Judge Grimm substantial airtime to develop ideas. However, follow-ups are often clarifying rather than challenging. The hosts do probe (e.g., 'Do you need a rule?' and 'Why has this not been addressed?'), but rarely push back on assumptions or press for tighter evidence. The conversation meanders pleasantly but lacks sharp editorial discipline typical of top-tier podcasts.

“Well, Judge, do you need a rule? I mean, the the situation you talked about, which was really interesting, it's not something that I had considered.”

“Why has this not been addressed in the last dozen years?”