Law Firms Need To Get Lawyers Off ChatGPT And Onto Artificial Intelligence That The Firm Can Monitor

Technology

Lawyers playing with artificial intelligence isn’t quite like handing dynamite to a series of pandas — it might be called “embarrassing” — but it’s close. Everyone is fired up and worried after a lawyer filed a case full of fictional law spewed by ChatGPT. A federal judge took the opportunity to grant an unfounded grandstanding in the case.

The problem in that case wasn’t the AI, it was the lawyer who failed to meet the professional obligations required when filing. Any For court research. It’s not ChatGPT’s fault that AI is the bargain basement retail solution. This is a new product slapped on top of another useful model. It’s like doing legal research through Yahoo while Lexis is sitting there.

But AI isn’t going anywhere. The robots may not be ready to consistently deliver what law firms need, but lawyers know what’s out there and use the tools available. And that means letting partners spit out the nightmare of turning loosely sensitive information into ChatGPT. Unless the organization gets ahead of the game and maintains a safe environment for AI testing.

That’s what Lega sees as an opportunity. Founded by former Ryan Court president and COO Christian Lang, Lega provides a means for law firms to securely test AI models, deploy API-driven applications for attorneys, and develop firm-wide policies for AI. Bring the use of AI under the corporate roof.

“Companies need to start,” Lang said, when it comes to integrating AI into the practice of law. “And it’s hard to know how to start.”

In the Lega environment, organizations have a sandbox to test future generative AI offerings (yes, there are others besides GPT), analyze results, and map out ways to introduce the technology into their workflows. On the latter point, Lega announced a partnership with Betty Blocks, a codeless app developer. Depending on the complexity of the request, an organization can spin up a new application in less than five minutes.

That is the long-term hope of the Lega model. The sandbox is good and except for the enterprise-scale opportunities of this sort of thing is to use the wisdom of the attorney crowd. “If a user finds a good question… you’re in. It’s there in the audit trail,” Lang explained. “Take this lesson and structure it and promote it to everyone.”

Within 90 days of its founding, Lega was already operating at Womble Bond Dickinson. The company’s Chief Innovation and Innovation Officer, Bill Koch, said in announcing the launch of Lega:

Lega not only provides enterprise defenses that allow us to feel confident in exploring LMMs and their use cases, but also provides critical tools for our professionals to quickly develop the lessons learned and designed to use these models. So that all of our users across the organization can benefit from the findings.

For all the legal faults of AI, the magnitude of this technology’s contribution to the provision 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 paws on the pandas. And the best way is to observe them and learn from their successes… and their mistakes.

Joe Patrice is a senior editor at Beyond Law and a contributor to Thinking Like a Lawyer. Feel free to email us with any tips, questions, or comments. Follow him. Twitter If you’re interested in law, politics and wholesome college sports news. Joe also serves as Managing Director in RPN Executive Search.