Attorneys playing with artificial intelligence isn’t exactly like handing dynamite to a bunch of pandas — perhaps fittingly called an “embarrassment” — but it’s close. Everyone got all hot and bothered after a lawyer submitted a filing teeming with fictional caselaw spit out by ChatGPT. A federal judge even took the opportunity to do some baseless grandstanding about it.
The problem in that instance wasn’t AI, it was a lawyer failing to meet the professional obligations required when submitting any research to a court. It’s also not ChatGPT’s fault that it’s the bargain basement retail solution of AI. This is a novelty product slapped atop an otherwise useful model. It’s like conducting legal research through Yahoo when Lexis is just sitting right there!
But AI isn’t going anywhere. The robots may not be ready to consistently deliver what law firms want, but lawyers know what’s out there and they’re going to use the tools available. And that means turning associates loose clacking potentially sensitive information into ChatGPT so it can spit out hallucinations. Unless the firm gets ahead of the game and maintains a secure environment for AI experimentation.
That’s where Lega sees opportunity. Founded by former Reynen Court president and COO Christian Lang, Lega offers law firms a mechanism to securely experiment with AI models, build API-driven applications for attorneys to deploy, and set firm-wide policies for AI. Bring the use of AI under the firm’s roof.
When it comes to integrating AI into the practice of law, “firms need to get started,” Lang said. “And it’s hard to know how to get started.”
In Lega’s environment, firms have a sandbox to test the up-and-coming generative AI offerings (yes, there are others besides GPT), analyze results, and figure ways to introduce the technology to the workflow. On the last point, Lega announced a partnership with Betty Blocks, the no-code app builder. Depending on the complexity of the ask, a firm can spin up a new application in less than five minutes.
That’s the long-term promise of the Lega model. Sandboxes are cool and all but the enterprise-scale opportunities of something like this is in leveraging the wisdom of the attorney crowd. “If a user figures out a cool prompt… you’ve got it. It’s there in the audit log,” Lang explained. “Take that learning and configure it and scale it to everyone else.”
Within 90 days of inception, Lega was already up and running at Womble Bond Dickinson. Bill Koch, the firm’s Chief Knowledge & Innovation Officer, noted in Lega’s launch announcement:
Not only does Lega provide us with the enterprise guardrails we need to feel confident exploring LLMs and their use cases, it provides a powerful set of tools to capture critical learning about where our professionals are inspired to use these models and then quickly scale the lessons learned throughout the enterprise so all of our users can benefit from breakthroughs.
For all the snickering about AI’s legal missteps, the magnitude of this technology’s contribution to the delivery of legal services should rival the introduction of computers and the internet.
Until then, law firms need to figure out how to keep all the limbs on the pandas. And the best way is to keep an eye on them and learn from their successes… and mistakes.
Joe Patrice is a senior editor at Above the Law and co-host of Thinking Like A Lawyer. Feel free to email any tips, questions, or comments. Follow him on Twitter if you’re interested in law, politics, and a healthy dose of college sports news. Joe also serves as a Managing Director at RPN Executive Search.