Description

If someone is going to make #AppellateTikTok a thing, it’s John Stobart. John is a certified appellate specialist who shares practical guidance on responsible AI use for lawyers through short videos on TikTok and at stobart.ai. John walks Tim and Jeff through how he actually uses Claude in an appellate practice, and where he draws the line between using AI and delegating to it.

We cover concrete workflows: how John built a verified case law corpus and a five-year brief bank to control what sources the AI can draw from, how he uses the Claude Word plugin's /check doc command as a final proofreading pass before filing, and why he still writes every argument section himself. (Well, the fact section is pretty dry stuff, so you could be forgiven for letting AI pick up the pen there.)

We also talk about California's newly signed SB 574, the first state statute regulating lawyers' use of generative AI, and what its verification and disclosure requirements mean in practice for litigators who do and do not use AI.

We cover:

• How to get started with AI without compromising the quality of your work product
• Building a brief bank and deposition corpus as verified sources for AI-assisted drafting
• Practical safeguards against hallucinated citations, including a Westlaw-to-Claude feedback loop
• What SB 574 requires of California litigators starting January 1
• Whether AI arbitration could ever be appropriate for wage-and-hour disputes

If you found this episode useful, please recommend the California Appellate Law Podcast to a colleague.