legal education
Is the unauthorized use of artificial intelligence in law schools a violation of the honor code?
“We’ve certainly had issues with students using generative AI in ways that faculty don’t think are appropriate,” said Benjamin Barros, dean of the law school at Stetson University in Florida. “I don’t want my students to accidentally find themselves in an honor roll situation.” (Image from Shutterstock)
As generative artificial intelligence becomes increasingly available and accepted into student workflows, some law schools are wondering whether misuse of AI should be a violation of honor codes. This violation can trip up aspiring attorneys during the character and fitness portion of the bar licensing process.
The fraudulent use of AI, which is considered a form of plagiarism, has raised questions about how best to deal with potential academic misconduct at Florida’s Stetson University School of Law.
“There are definitely issues with students using generative AI in ways that faculty don’t think are appropriate,” said Benjamin Barros, dean of the Florida School of Law. “We don’t want our students to accidentally find themselves in an honor roll situation.”
“We don’t have to tackle the honor code head on, right? And that’s a real challenge right now as we’re all trying to figure this out,” he added.
Before being admitted to the bar, applicants must pass the character and fitness portion of the bar admissions process, and the most common burden of proof for applicants is having to show clear and convincing evidence of good moral character, said David L. Hudson Jr., an associate professor at Belmont University College of Law in Tennessee.
“An allegation of academic misconduct is one of the major red flags,” says Hudson, who represents bar applicants in character and fitness cases and frequently writes about legal ethics for publications such as the ABA Journal. “This is one of the most serious hurdles to overcome.”
Richard M. Trachok II, chairman of the Nevada Bar Examining Board, said the Nevada Bar Examining Board treats honor code violations very seriously and candidates found guilty of violations may not be allowed to take the bar exam.
He says the circumstances of each case will be carefully considered and the outcome will depend on what specific instructions the professor gave the students.
“If that instruction says, ‘Do not use AI,’ and someone ignores that instruction, that’s clear fraud,” Trachok added. “That’s pretty tough.”
He said the disciplinary hearing process will mirror the process for any plagiarism crime, noting that the Nevada Board of Bar Examiners has not yet encountered such a case.
“We will refer the matter to the Character and Fitness Commission, which will conduct the investigation,” Trachok said. “They will make a judgment and ultimately I will be the one deciding on the sanctions.”
If candidates protest the ruling, they could appeal to the Nevada Supreme Court, he added.
Daniel W. Rinna Jr., Northwestern University Pritzker School of Law: “If we have a policy that says we can’t use generative AI to do substantive course work, and a student does a Google search, should they look away?” he says. “At the top of the page is generative AI.”
lack of clarity
Daniel W. Rinna Jr., a senior lecturer and director of law and technology initiatives at Northwestern University’s Pritzker School of Law in Illinois, said the problem stems from unclear AI policies within law schools and universities.
These cases “show why these policies are problematic,” says Rinna in the 2018 Journal Legal Rebel.
Most of the policies Rinna saw in law school do not draw clear lines between what is acceptable and what is not.
“If we have a policy that says generative AI can’t be used to do substantive course work, and a student does a Google search, should they look away?” he says. “At the top of the page is generative AI.”
The University of California, Berkeley School of Law’s policy specifically states that generative AI may not be used for any purpose that would be considered plagiarism if a human were involved, and “may not be used for any purpose in any exam context.”
The use of certain forms of AI is a violation of the honor code, said Chris Hoefnagle, professor and dean of the school’s Berkeley Law and Technology Center.
“When students summarize using generative AI, the system introduces a fake case and that case is put on the exam. We are looking at using generative AI,” Hoofnagle says.
Paul Rose, dean of the Case Western Reserve University School of Law in Ohio, said policies are being fine-tuned as AI advances.
“We try to be as clear as possible about when AI can and cannot be used,” Rose says.
Even if a student is found to have used it incorrectly, there is still the possibility of disciplinary action being reported to state court, he added.
Barros said Stetson University School of Law faculty members also have ongoing conversations about the use of AI.
“Things are changing rapidly,” he says. “We don’t want to set up what amounts to a perjury trap for our students, right? We want to be transparent and have good communication with our students.”
Rinna also points out that the tools currently available to detect the use of AI are unreliable, with some tools biased toward non-native English speakers, for example.
“We don’t have a good way to police this,” Rinna says. “What happens if someone is wrongly accused or makes an innocent mistake? This should really force law schools to rethink what we are trying to accomplish with these policies and whether we are doing more harm than good.”
Kelly Testi of the Association of American Law Schools: “No one wants a student to trip over a bar because of something they did inadvertently. That’s traumatic,” Testi said. “AI is integrated and everywhere. We are better off empowering people.” (Photo by Josef Karinko)
According to Kelly Testi, executive director and chief executive officer of the American Association of Law Schools, the solution includes a clear AI policy as well as robust ethics training for law students to use AI before entering the workplace, where ease of use of the tools is expected.
“We saw a real seismic shift in the last year in terms of schools getting serious, making things clear for their students, and recognizing AI as a tool that needs to be understood and used appropriately,” says Testy of the 2022 Journal Legal Rebel.
Law school education must include critical thinking, along with clear communication between students and professors, administrators note.
“Nobody wants a student to accidentally trip up at a bar. That’s traumatic,” Testi says. “AI is integrated and everywhere. We are better off empowering people.”
Write a letter to the editor, share a story tip or update, or report an error.
