The Rise of the Robot Lawyer? DoNotPay’s Legal AI Faces Several Challenges

This article was originally published March 10, 2023.

The first few months of 2023 have witnessed an unfolding drama over a clash between the worlds of tech startups and the legal profession. While the initial promise that the new year would bring us the first ever “robot lawyer” was retracted before January was over, the excitement surrounding the legal-tech frontier shows no sign of diminishing.

Enter (And Exit) the Robot Lawyer

Battle lines were drawn early this year over the consumer advocacy app DoNotPay, whose founder and CEO Joshua Browder announced on January 5th that an artificial intelligence (AI) legal assistant created by DoNotPay, touted as “the world’s first robot lawyer”, would argue a case in an undisclosed traffic court sometime in February. According to the announcement, this unprecedented event was to be achieved by having the unnamed litigant repeat verbatim the arguments generated by the DoNotPay app. The AI would listen to the court hearing in real time and dictate to the litigant, who would be sporting some form of recording device such as high-tech smart glasses plus an earpiece, what to argue to the court. DoNotPay reportedly agreed to cover any fines, should the litigant lose the case.

Just days after this announcement, Browder upped the ante, tweeting on January 8th that:

DoNotPay will pay any lawyer or person $1,000,000 with an upcoming case in front of the United States Supreme Court to wear AirPods and let our robot lawyer argue the case by repeating exactly what it says (1/2)

He continued,

We have upcoming cases in municipal (traffic) court next month. But the haters will say “traffic court is too simple for GPT.” So we are making this serious offer, contingent on us coming to a formal agreement and all rules being followed. Please contact me if interested!

This attention-grabbing announcement naturally sparked some cheers and a significant amount of backlash, with lawyers and non-lawyers alike voicing skepticism that such a move would ever be allowed in the Supreme Court. Much of the criticism focused on the logistical problem of the Court’s restrictions on physically bringing any electronic devices into the courtroom. A few commenters also noted that any lawyer slated to argue before the Court would be risking contempt and/or violating their ethical duties, putting their license on the line, should they take Browder up on his offer (and to our knowledge, nobody in the Supreme Court Bar did).

At the time, Browder, whose Twitter profile tagline quotes the BBC dubbing him the ‘Robin Hood of the Internet’, was undeterred by these concerns, even as they applied to the lower-stakes arena of traffic court. On January 20th Browder confirmed his intent to forge ahead with the game-changing move of having AI argue before a judge on February 22nd.

However, by January 25th, Browder tweeted that after receiving “threats” from unspecified “State Bar prosecutors” indicating that he could be jailed for six months if he were to proceed with the scheme, DoNotPay would be “postponing” the court case and “sticking to consumer rights” issues such as “lowering medical bills, cancelling subscriptions, disputing credit reports, among other things.” As revealed in later interviews, Browder was notified by state bars, and advised by his (human) counsel, that if he proceeded with the traffic court plans, he could be facing potential criminal charges for the unauthorized practice of law.

So, Browder continued, DoNotPay would be backing away from “non-consumer rights legal products (e.g defamation demand letters, divorce agreements and others)” and removing these legal products from DoNotPay “effective immediately, to focus solely on consumer rights.” However, as of the date of this post, the company’s website still prominently calls the app “The World’s First Robot Lawyer.” In addition to the routine consumer activities such as cancelling gym memberships, closing bank accounts, or converting travel points, the website still lists among its services: analyzing terms of service, annulling marriages, breach of contract, copyright protection, defamation demand letters, revocable and irrevocable trusts, trademark registration, and suing companies in small claims court. The full scope of such services is unclear. So long as such legal products continue to be offered (and to the extent they are marketed with the dramatic flair that has come to characterize this company’s social media activity), however, ethical, statutory, and common law liability issues continue to be implicated.

The Legal and Professional Barriers

In his announcement backing away from the plan to send the DoNotPay robot lawyer to court, Browder cited state bar threats to prosecute him for the unauthorized practice of law (UPL), a misdemeanor offense that in many states may rise to a felony if certain aggravating factors are present.

In addition to the UPL concerns, there are several other issues raised by DoNotPay’s proposal to allow AI to argue a court case in real time, especially if an attorney is involved either in the development or use of the technology. These issues include competence (See ABA Model Rule 1.1), the exercise of independent professional judgment (Model Rule 2.1) , frivolous claims and candor to the tribunal (Model Rules 3.1 and 3.3), conflicts of interest (Model Rules 1.7 and 1.8), and confidentiality (Model Rule 1.6). The legal malpractice risk is also prominent. As for a pro se party, offering to pay a litigant to bring or pursue a claim as a PR stunt raises such ancient common law concepts as champerty, maintenance, and barratry, each of which are still recognized offenses or claims in several jurisdictions aiming to combat frivolous litigation.

In this multi-part series, we will discuss each of these ethical, statutory, and common law issues in depth as we explore how the regulation of the legal profession can be applied in the LegalTech industry.

What About the Consumer?

Of course, there is tremendous value and societal benefit to having cheap, easy-to-use, automated tools helping everyday consumers avoid excessive costs that can be a pain to cut. I myself recently had to go through four rounds of email exchanges just to cancel an automatically renewing antivirus software subscription. An AI assistant would have saved me time, and more than a few exasperated sighs.

It is also true that wider access to quality legal services, which can be prohibitively expensive for most people, remains an ongoing issue for the industry, and finding ways in which technology can reduce this burden is and ought to be a perennial pursuit. But where should we draw the line between AI performing the kind of routine form-filling that has been recognized as distinct from offering legal advice and tasks that should be undertaken by a duly licensed professional in order to protect consumers of legal services?

As AI continues to develop increasingly sophisticated research, issue-spotting and communication capabilities, it will likely become not only a useful, but even a necessary tool for lawyers in order to remain competitive in the legal services industry. However, as long as the reasons for having licensing structures and ethical obligations remain relevant, the return of the “robot lawyer” to replace the human one seems unlikely to happen anytime soon.