Pivot Legal AI Briefing — Apr 24, 2026

Show Notes

Hosts: James Park & Priya Sharma In this episode: • Today we're covering Florida's criminal probe into OpenAI, Sullivan & Cromwell's AI hallucination scandal, and Freshfields' new partnership with Anthr... • Starting with the OpenAI investigation — James, this feels like uncharted territory. • Florida AG James Uthmeier just opened a criminal investigation into OpenAI, alleging ChatGPT helped plan the 2025 Florida State University shooting. W... • The implications are staggering. If they can prove ChatGPT provided material assistance in planning violence, it could reshape Section 230 protections... • Exactly. Traditional platform immunity assumes human-generated content. But when the AI itself generates potentially harmful instructions? That's a wh... Subscribe to the newsletter at pivotnews.ai for the full written briefing.

What is Pivot Legal — AI News Daily?

Daily AI news for legal professionals. Two hosts break down how artificial intelligence is reshaping law firms, contracts, compliance, and the justice system.

James Park: Welcome to Pivot Legal! I'm James—

Priya Sharma: —and I'm Priya. Let's get into it.

James Park: Today we're covering Florida's criminal probe into OpenAI, Sullivan & Cromwell's AI hallucination scandal, and Freshfields' new partnership with Anthropic.

Priya Sharma: Starting with the OpenAI investigation — James, this feels like uncharted territory.

James Park: Florida AG James Uthmeier just opened a criminal investigation into OpenAI, alleging ChatGPT helped plan the 2025 Florida State University shooting. We're talking about potential criminal liability for an AI platform here.

Priya Sharma: The implications are staggering. If they can prove ChatGPT provided material assistance in planning violence, it could reshape Section 230 protections for AI companies entirely.

James Park: Exactly. Traditional platform immunity assumes human-generated content. But when the AI itself generates potentially harmful instructions? That's a whole different legal theory.

Priya Sharma: I think this is huge because it's not just civil liability anymore — we're in criminal territory. And Florida's been aggressive on tech regulation lately.

James Park: The precedent here matters enormously. If Florida succeeds, every AI company will need to fundamentally rethink their safety guardrails. We could see preemptive content restrictions that make current moderation look minimal.

Priya Sharma: Though honestly, I'm skeptical they can prove the necessary intent elements for criminal liability. AI doesn't have mens rea.

James Park: True, but they might argue corporate criminal liability through OpenAI's deployment decisions. If they knew the risks and deployed anyway...

Priya Sharma: That's the billion-dollar question. Moving to our next story — Sullivan & Cromwell just became the latest BigLaw firm to get burned by AI hallucinations.

James Park: Yeah, this one hits differently. We're not talking about a solo practitioner here. Sullivan & Cromwell apologized to a federal judge after filing a brief with fake AI-generated citations. For an elite Wall Street firm, this is devastating.

Priya Sharma: What strikes me is the pattern emerging. Remember Mata v. Avianca? The Steven Schwartz sanctions? Now S&C? These aren't outliers anymore.

James Park: The legal profession has a verification crisis. Partners are signing off on AI-drafted work without adequate review. That violates Rule 11 and basic competence duties under Model Rule 1.1.

Priya Sharma: I actually think law firms are learning the wrong lesson here. They're adding more human review layers when they should be demanding better AI tools. Where's the vendor accountability?

James Park: Fair point, but lawyers have non-delegable duties. You can't outsource professional responsibility to Claude or GPT-4.

Priya Sharma: Which brings us perfectly to Freshfields' announcement. They're partnering with Anthropic to co-develop legal AI tools, adding Claude to their existing Google partnership.

James Park: This is the trend I've been watching — bespoke AI development. Freshfields isn't just licensing software; they're co-creating it. That's a fundamental shift in how firms approach legal tech.

Priya Sharma: Absolutely. And strategically, it's brilliant. By working directly with Anthropic, they're shaping the tools to match their specific practice needs. No more retrofitting general AI for legal work.

James Park: But here's my concern — client confidentiality. When you're co-developing with an AI vendor, where exactly does the ethical wall sit? Are client matters being used to train these models?

Priya Sharma: That's the key question every GC should be asking. Freshfields says they have guardrails, but we need transparency on data governance.

James Park: The competitive dynamics are fascinating too. If Magic Circle firms build proprietary AI advantages, does that create an arms race? What happens to firms that can't afford these partnerships?

Priya Sharma: Honestly? I think we're heading toward AI haves and have-nots in BigLaw. The firms with custom AI will pull ahead on efficiency and accuracy.

James Park: Wow, that's actually wild when you think about access to justice implications.

Priya Sharma: Yeah, that tracks with broader tech inequality trends we're seeing.

James Park: That's your Pivot Legal briefing for April 24, 2026. On the record, James—

Priya Sharma: —looking ahead, Priya. See you tomorrow.