Free CLE Eligible · Plaintiff Med-Mal Only

Cited Case Law Only — Auditing Your AI Stack for Malpractice Risk

I audited 50 AI-generated demand letters. 34 had fabricated citations. Learn the 3-minute audit methodology before your next one goes out the door.

Tuesday, May 13, 2026 12:30 PM ET · 45 min Online (Zoom) Free

Register — takes 20 seconds

✓ You're in — check your email

We just emailed you a calendar invite with the Zoom link. See you Tuesday May 13 at 12:30 PM ET.

Didn't get an email? Check spam or email [email protected].

What you'll learn

The Mata v. Avianca architecture problem. Why every AI legal tool hallucinates citations by default — and the 3 architectural choices that prevent it.
The 50-letter audit. Real output from Supio, EvenUp, ChatGPT, Claude Desktop. Where each tool fails. What the failure pattern looks like.
The 3-minute audit checklist. How to run this on any AI tool you're evaluating (or already using) in under 3 minutes. Paste-ready spreadsheet template included.
CourtListener + PubMed grounding. The two free retrieval sources that eliminate hallucination risk for any legal AI — with code examples.
Live demo. Watch MedLegal AI analyze a real NJ cardiology depo transcript, surface impeachment literature from PubMed, and generate a demand letter with real CourtListener citations — in front of you.

Agenda

0-5 min

Why you can't just "tell the model not to hallucinate"

The architectural problem in every current legal AI tool. Named examples.

5-15 min

The 50-letter audit: what I found

Specific examples of fabricated citations. How they slipped past attorneys. Which tools fail worst.

15-25 min

The 3-minute audit methodology

Paste-ready spreadsheet template. Three categories of hallucination. How to spot each.

25-40 min

Live demo: CourtListener + PubMed grounding

Watch MedLegal AI generate a demand letter with real citations. Real case: Crossin v. Shackelford, cardiology misdiagnosis.

40-45 min

Q&A

Open questions. Paste your own AI-generated output and I'll audit it live.

Who should attend

This webinar is for plaintiff medical-malpractice attorneys — solo practitioners, small-firm partners, and firm-level legal ops leads who've either already deployed AI tools or are evaluating them.

If you're at a firm running ChatGPT, Claude, Supio, EvenUp, CoCounsel, or Harvey on any plaintiff-side work product, this audit will show you exactly where the risk is hiding.

Speaker

John Mahoney

Founder, MedLegal AI

John is the founder of Medicolegal Intelligence LLC (Delaware, 2026). He spent 18 months building 23 AI tools specifically for plaintiff medical-malpractice attorneys, with an architectural commitment to zero hallucinated citations. Prior to MedLegal AI he spent 12 years in software engineering with a focus on natural-language systems. He writes the Deposition Intelligence newsletter.

Attendee perks

6-month free Pro trial of MedLegal AI (normally $249/mo) — emailed to all registrants within 24 hours of the webinar.
Audit spreadsheet template — the same 3-minute audit methodology, paste-ready.
CLE certificate — pending bar approval in NJ/NY/PA/FL/CA (check your state).
Recording sent to all registrants within 48 hours, even if you can't attend live.