This article looks at a high-profile mix-up involving Sullivan & Cromwell, where AI-generated errors slipped into a federal bankruptcy court filing. Through Max Raskin, co-founder of Uris Acquisitions and an NYU Law fellow, we get a closer look at what happened and why it matters for the legal field.
Why does this matter? Well, the whole thing highlights the tension between innovation and accountability. AI’s already everywhere in law, so the real conversation should probably be about using it responsibly, not shutting it down entirely.
AI in legal practice: lessons from Sullivan & Cromwell
The main event? A firm filed a court document that was packed with mistakes from an AI tool. It was embarrassing, sure, but it didn’t mean the tech itself was a lost cause.
Even top lawyers lean on AI for drafting, research, and analysis. Max Raskin points out that if elite practitioners get AI help, clients with fewer resources deserve the same shot.
Honestly, the reach of AI in law is only growing. Ignoring it just doesn’t seem realistic anymore.
The incident and its implications
Raskin believes the blunder should push law firms to get better at using these tools, not to ban them outright. Technology isn’t really the enemy here—the real challenge is how we use these tools, double-check their work, and catch mistakes before they get out.
He thinks courts should let AI chatbots and assistants in, as long as there’s strict oversight. That shifts the debate away from “should we or shouldn’t we” toward “how do we use this stuff safely and transparently?”
Why Raskin argues for access and safeguards
Raskin says everyone should get access to AI-powered tools, since legal practice has already changed so much thanks to modern tech. If big firms use these resources, clients should get the benefits too—without losing out on accuracy or fairness.
Let’s face it: using AI in law isn’t really optional anymore. It’s about keeping things efficient and consistent, but only if we pair it with solid safeguards.
Key takeaways for practitioners
- Transparency about how AI helps with drafting or decisions lets clients know the risks and how much they can trust automated results.
- Oversight is key—someone needs to check AI-generated content before it gets filed or used in court.
- Accountability means human editors have to take final responsibility for whatever gets submitted.
- Error mitigation—like validation checks and audit trails—helps catch AI-driven mistakes before they become a problem.
Regulation by design: transparency, oversight, and error mitigation
The bigger idea? Guide regulation toward responsible use, not blanket bans. A framework built on transparency and oversight can make space for AI chatbots in court, while still keeping standards high for accuracy and fairness.
In practice, that means setting up rules for checking AI output, labeling when work is AI-assisted, and always having a human involved at key steps.
Practical safeguards in a modern practice
- Set up verification protocols for AI-generated work, and require a human to review everything before it goes to court.
- Keep audit trails that show what the AI did, how it reasoned things out, and what the final output was, so there’s a record.
- Use species-specific safeguards to fit the type of law—bankruptcy, tax, litigation—since each has its own risks.
- Make sure lawyers get training and education on AI’s limits, biases, and where it tends to mess up.
Conclusion: a path to a modern, accountable legal tech ecosystem
The Sullivan & Cromwell incident isn’t really a reason to fear AI. Instead, it’s a wake-up call for a smarter, safer approach to using AI tools in law.
Law firms can modernize their work and boost efficiency by choosing controlled AI use. But let’s be real—clear transparency, careful oversight, and solid error checks have to come first.
AI should add value, not replace the judgment that makes legal work trustworthy. Human expertise has to stay at the center of it all.
Here is the source article for this story: Let the chatbots practice law