Written by Joel Engardio for Everlaw

News of the Robot Lawyer Is Premature

Robot lawyers are making headlines everywhere from USA Today to the Washington Post thanks to Stanford University student Joshua Browder, who has created a chatbot he claims is powering “the first case of a fully automated lawsuit.”

“You can sue Equifax for $25,000 by pressing a button,” Browder tweeted.

Browder wants to empower consumers affected by the credit agency’s massive security breach, which compromised the personal data of up to 143 million people in the U.S. — including his own. His chatbot promises a free and easy way to sue Equifax without having to hire a lawyer.

Browder originally created his DoNotPay chatbot to help people fight parking tickets. The bot uses what Browder describes as Artificial Intelligence (AI) to find relevant legal forms based on answers to questions like, “What is your zip code?” Now Browder has unleashed the chatbot to automatically fill out the forms required for suing Equifax in small claims court.

But is this robot actually a lawyer? Writer and lawyer Ephrat Livini unequivocally says “no” in a Quartz essay cleverly formatted as a legal brief — complete with sections for standing, facts and a motion to dismiss.

Livini points out that a chatbot “cannot visit you in jail, represent you in court, sign a document, convince the opposition, cajole a judge, charm a prosecutor, pay your bail bondsperson, negotiate a resolution, or be held liable criminally or civilly for ineffective assistance.”

For all these reasons, talk of robot lawyers is obviously premature.

What about Browder’s AI claims? At best, his chatbot can find and fill out legal forms. That’s a convenience. But it isn’t AI or a “fully automated lawsuit.” Much more must happen to take a complex case to court and win. A seasoned human lawyer is still needed to appear in court and make a convincing argument to prove wrongdoing or defend innocence.

Even if Browder’s chatbot used actual machine learning, it still wouldn’t replace human lawyers.

“The practice of law is a bewildering collision of contexts: language, business operations, societal structure and mores, law, slang, implication. It’s subjective and ill-defined, logical and emotional,” Everlaw CEO A.J. Shankar told Bloomberg Law. “Gray areas abound. It’s the sort of environment where humans excel — and AI shows its limitations.”

To be sure, machine learning can be a huge help in litigation. Staying up all night digging through boxes of legal files to find a needle-in-the-haystack memo is hardly a good use of a lawyer’s time. And the deluge of data search can put firms and government agencies at a disadvantage when they lack the resources to pay a team of lawyers to sift through thousands of paper files.

Predictive coding, or TAR, systems use machine learning to quickly identify documents that are relevant to a case, and weed out out superfluous documents so lawyers can focus on finding the smoking guns.

But this isn’t about robots taking jobs from lawyers—much less becoming lawyers themselves. This is about using technology to help humans practice better law.