By Dajuan Davis & Martin Dolan
By now, most attorneys have received a form of solicitation from a tech company urging them to sign-up for an AI Chatbot program. Previously dismissed as spam, digital advertisements with flashy headlines announcing the unbelievable benefits of AI Chatbots are now becoming difficult to ignore. Advertised as an innovative technology that assists in advertising legal services, drafting documents, and developing creative legal strategies, AI Chatbots are becoming tools that attorneys are now using to get ahead of their competitors, and to make their firms more profitable. The introduction of AI Chatbots in the legal field may benefit attorneys, firms, legal consumers. However, attorneys must exercise due care when employing this new technology. Failing to do so may result in the attorney running afoul of the rules of professional conduct and facing severe consequences.
Innovation in the legal field
Innovations in the legal field is nothing new. Attorneys and other legal professionals across the United States previously conducted legal research by finding their nearest law library, pouring over books, and tirelessly researching caselaw. As some lawyers may painfully recall, this pre-1985 manual way of conducting legal research was grueling and time-consuming. However, it was necessary in order for attorneys to provide diligent and competent legal services to their clients. In the mid-1970s, sites such as Westlaw, and LexisNexis began compiling American caselaw into their online platforms. The emergence of these “online libraries” was a revolutionary moment for the legal profession. There was no longer a need to travel or handle physical books. Attorneys could simply log on to one of these sites to conduct their research, saving an immense amount of time and effort, which benefited their clients greatly. Nevertheless, this evolution took time, as most law schools had only a handful of computers at best. The introduction of online of online libraries was also met with skepticism. As Westlaw, and LexisNexis was being introduced into law schools across the country, legal professionals voiced concerns over the potential pitfalls of using the new revolutionary tools. Law Students were taught to not solely rely on the information located on these sites. Instead, students were expected to double check their work when utilizing online libraries and to continue to use the old method of legal research. Now decades later, tools like Westlaw and LexisNexis are the standard way in which attorneys conduct legal research, and online libraries are an integral part of attorneys’ day-to-day lives.
As innovation in technology continues to advance, the legal profession seems to be on the precipice of another moment of great change. With the rise of new technologies and artificial intelligence, legal professionals are now able to simply input a legal question or task into a AI Chatbot program, and the program will provide the answer, research, and or draft a document for the user at a fraction of the time it would have taken the user to do it themselves.
Developers have been working on these forms of Artificial Intelligence for years. Now, AI Chatbots have the ability to understand human language, solve complex problems, generate human-like written responses, and draft legal documents. AI Chatbots have proven themselves intelligent enough to pass a U.S medical and licensing exam, Wharton Business School exams, and law school exams. These program have also demonstrated the ability to write full length essays on topics that a user inputs. Other programs have been specifically developed to complete legal tasks. For instance, some AI programs have advanced capabilities in drafting legal instruments such as contracts. While considering the entire contract, programs can suggest new language or points for negotiating. These programs are now being used by lawyers to ensure they do not forget to include pertinent provisions while drafting a contract.
Other AI programs were developed specifically for the legal profession to assist lawyers more broadly. With the goal of automating grueling aspects of the legal profession, developers have created AI programs that already can perform legal tasks such as research, compliance, and drafting documents for litigation. One program, Harvey AI, is already being used by Allen & Overy, the seventh largest law firm in the world. Over 3,500 Allen & Overy attorneys in 43 offices have used the platform to assist them in their day-to-day legal tasks. According to the firm, the program has delivered unprecedented efficiency and intelligence, and they have seen some “amazing results” while using Harvey. David Wakeling, Head of the Markets Innovation Group, at Allen & Overy believes that AI Chatbots like Harvey are “game changers” that can “transform the legal industry.” In March 2023, another major firm, PricewaterhouseCoopers LLP, subsequently announced the introduction of an AI Chatbot service to its firm. Believing that its lawyers will benefit from the program’s assistance in contract analysis, regulatory compliance work, and legal consulting services, the firm has provided program access to over 4,000 of its legal professionals.
AI Chatbots and Ethical Pitfalls
The use of AI Chatbots in the legal field will only become more popular in the coming years. Despite the unquestionable benefits to attorneys, their firms, and clients, the use of this technology is not without ethical pitfalls. Attorneys must abide by the rules of professional conduct and exercise due care while using AI Chatbots to perform legal tasks or to interact with clients. It is not difficult to imagine a future where a busy attorneys may be tempted to simply rely on an AI Chatbot’s inaccurate research, or a document the AI creates. An attorney might submit an AI created document fraught with errors to the court, or dishonestly present an AI created document to the court as their own work-product. These behaviors may run afoul of the rules of professional conduct.
At the heart of the legal profession is an attorney’s obligation of good faith to provide diligent and competent legal representation to legal consumers in both preparation and presentation of their client’s case. Melissa Smart, of the Attorney Registration and Disciplinary Commission of the Supreme Court of Illinois articulates that “part of an attorney’s responsibility when employing nonlawyers, is to play the role of a filter.” When nonlawyers such as paralegals, research assistants or law clerks draft documents for an attorney or a firm, attorneys must diligently look over the information, language, and details of the documents to make sure they are correct, and professional. Carelessly relying on the work or research output of any non-lawyer may not only constitute misconduct per the rules of professional conduct but can also result in a malpractice action against the attorney. This is true in an instance where an attorney carelessly relies on the results of an AI Chatbot as well. Despite the intelligence of AI Chatbot programs, these programs are also nonlawyers that have periodically produced inaccurate and misleading results. Attorneys should diligently review the outputs of these programs to make sure the information in the program’s output is correct.
Other rules of professional conduct may also be implicated while using AI Chatbots, depending on the attorney’s use of the platforms. For instance, generally, attorneys cannot assist another “person” in the unauthorized practice of law in violation of the rules governing professional conduct in that person’s jurisdiction. The definition of “practicing law” is determined by law and varies from one jurisdiction to another. However, practicing law generally includes providing legal advice and or legal services to legal consumers. At the time of this article’s writing, it is unclear whether AI are considered “persons” in the rules of professional conduct. Joshua Browder, the CEO of tech company DoNotPay, was set to employ an AI-powered robot to argue for the defense in a traffic ticket case in California. This would have been the first time AI has ever argued in an American courtroom. However, the CEO abandoned the experiment, claiming that he was threatened by multiple state bars including the California State Bar. Browder allegedly faced criminal investigation and possible prosecution for his experiment. The unauthorized practice of law is a misdemeanor in some states and punishable up to six months in jail. Although the rules of professional conduct are still evolving to adequately account for the role of AI in the legal profession, attorneys may face severe disciplinary measures and criminal punishment if they assist AI in providing legal advice, or legal services to clients or prospective clients.
Attorneys should also consider the implications of removing the human element out of aspects of their legal services. The role of the attorney in an attorney-client relationship is important. Although AI Chatbots have proven themselves quite intelligent, some lack the ability to truly understand things like hyperbole, sarcasm, and other such complexities of human communications. As of now, programs have only been trained to generate words based on human input. Thus, a program’s response may lack insight or nuance and could potentially lead to error or miscommunications. Potential malfunctions or other mistakes by the program are also possible, and this is concerning when dealing with confidential information. The rules of professional conduct generally dictate that attorneys cannot reveal information relating to the representation of a client without the client’s consent. If a program mistakenly disseminates a client’s privileged information to a third-party, the attorney may find themselves held accountable by the disciplinary board in their state.
Attorneys Must Conduct a Thoughtful Balance
When exploring these new revolutionary tools in the practice of law, it is important to remember that technologies can and have assisted us in our endeavor to provide more extensive and affordable legal services to legal consumers. New technologies should be encouraged and embraced carefully. “Attorneys should conduct a thoughtful balance between embracing these new AI technologies and maintaining their professional duties and responsibilities.”
In the near future, these programs may be seen as invaluable tools for attorneys and firms. Much like Westlaw and LexisNexis removed time consuming aspects out of legal research, AI Chatbots will cut down the time an attorney spends on researching, strategizing, communicating, and drafting documents. As the use of AI Chatbots in the legal field continues gain in popularity, one can easily see how an attorney may become tempted to carelessly rely on these programs in such a way that lands them in an ethical dilemma. We are entering into a brave new world, particularly from an ethical and regulatory perspective regarding these emerging technologies. Thus, it is important that attorneys lacking guidance on a particular issue regarding the appropriate utilization of AI chatbots, make an inquiry to the relevant disciplinary board in their state to ensure they are not running afoul of established rules of professional conduct. Additionally, attorneys and firms should have in place policies and procedures for those using these technologies to complete legal tasks and ensure that their malpractice insurance will cover any incidents arising from the use of AI Chatbot programs.
Citations
1 Lakshmi Varanasi, OpenAI just announced GPT-4, an updated chatbot that can pass everything from a bar exam to AP Biology. Here’s a list of difficult exams both AI versions have passed., Business Insider, (Mar. 21, 2023, 12:36 PM).
2 David Wakeling, A&O announces exclusive launch partnership with Harvey, Allen & Overy, (Feb. 15, 2023).
3 Sara Merken, PwC’s 4,000 legal staffers get AI assistant as law chatbots gain steam, Reuters, (Mar. 15, 2023, 1:31 PM).
4 Model R. Prof. Conduct 1.1 (ABA 2011). Model R. Prof. Conduct 1.3 (ABA 2011).
5 See In re Hessinger & Associates, 192 B.R. 211, (N.D. Cal. 1996) (applying California law) People v. Milner, 35 P.3d 670 (Colo. O.P.D.J. 2001) Sussman v. Grado, 746 N.Y.S.2d 548 (Dist. Ct. 2002).
6 Bobby Allyn, A robot was scheduled to argue in court, then came the jail threats, NPR, (Jan. 25, 2023, 6:05 PM).
7 Model R. Prof. Conduct 1.6 (ABA 2011).