Before you ask: No. The robot didn’t pass the bar exam, so it’s not a licensed lawyer. However, that is not a requirement for arguing a legal case. People represent themselves and hire paralegals in court proceedings all the time. It’s not a stretch for a judge to agree to hear a case from an AI. In fact, most judges would probably be very interested to see a machine-generated legal argument, especially one presented in real-time.
Several years ago, we covered how a machine-learning algorithm bested 20 human lawyers when analyzing risks in nondisclosure agreements. The AI tied the highest-scoring lawyer with 94 percent accuracy. At the time, we predicted that it probably wouldn’t go much further than that.
“Are lawyers at risk of being replaced? Probably not, at least not for things such as arguing case law,” yours truly posited. Well, now I stand corrected.
NewScientist reports that an AI will argue the first legal case ever in a court of law in February. The hearing won’t be anything too exciting. It’s just a routine speeding ticket, which is probably why the court agreed to allow the unprecedented counsel request. The defendant’s life is not hanging in the balance, just a relatively inexpensive fine.
Of course, there won’t be an Android walking around the courtroom addressing the judge and jury — AI has come a long way since 2018, but not that far. Instead, an iPhone will be in the defendant’s pocket. A phone equipped with an AI app and earpiece will provide the user with the appropriate responses to arguments during the hearing.
Do not be mistaken. The AI used in this case is not the same as the NDA-analyzing bot from 2018. Consumer advocate organization DoNotPay developed this algorithm to help users get out of fines, fees, and subscriptions. It can also aid in procuring refunds from businesses that want to make it appear you have no choice but to eat the costs of their mistakes.
“The DoNotPay app is the home of the world’s first robot lawyer,” it boasts. “Fight corporations, beat bureaucracy, and sue anyone at the press of a button.”
The company has not disclosed the court’s location or the defendant’s name to ensure a controlled environment for the experiment. The defendant has been instructed only to say what the bot tells him. Should the judge rule against the defendant, NoNotPay agreed to cover any fines and fees associated with the case.
DoNotPay’s “robot lawyer” started life as a chatbot in 2015, similar to the types that annoy you on tech support sites. You know, the ones that are dumb as rocks and can never satisfactorily answer a simple question. In 2020, DoNotPay incorporated more sophisticated AI that it hopes is good enough to argue an actual legal case in real-time successfully.
DoNotPay’s founder and CEO, Josh Browder, said he wants to use the case to make his AI accurate and honest.
“We’re trying to minimize our legal liability,” Browder told NewScientist. “And it’s not good if it actually twists facts and is too manipulative.”
Ultimately Browder would like to see his app become good enough to replace an attorney but at a substantially lower fee.
“It’s all about language, and that’s what lawyers charge hundreds or thousands of dollars an hour to do,” he said. “There’ll still be a lot of good lawyers out there who may be arguing in the European Court of Human Rights, but a lot of lawyers are just charging way too much money to copy and paste documents, and I think they will definitely be replaced, and they should be replaced.”
It should be fascinating to see how this case turns out. We’ll be looking forward to a verdict sometime after February.
Image credit: The People Speak!