Using generative artificial intelligence? Ethics guidance is available
Jennifer Wondracek compares generative artificial intelligence to a 1L summer associate.
“It knows just enough to be dangerous. It doesn’t realize yet that there isn’t always an answer. And it really wants to make you happy,” said Wondracek, a professor of legal research and writing at the Capital University Law School in Columbus, Ohio. “So if you treat AI like you treat your 1L summer associate, make sure you’re checking your work.”
Wondracek, who is also the director of Capital University’s law library, spoke Thursday during an ABA Techshow 2025 program titled, “Ethical Panel Considerations for Lawyers Using Artificial Intelligence.” She and her co-panelists shared their approaches to using generative AI in legal practice, as well as potential risks to keep in mind.
There are currently more than 30,000 AI tools, said Wondracek, citing a website called There’s an AI for That, which tracks and categorizes these tools. She suggested that lawyers stick with the “big names in legal research”—Lexis+ AI, Thomson Reuters (which owns Westlaw) and vLex Fastcase.
“The great thing about these systems is they are built by companies that understand lawyers,” Wondracek said. “So they tend to default to the privacy and ethical obligations that we need.”
Several recent ethics opinions provide guidance on how to use generative AI, including the Pennsylvania Bar Association and the Philadelphia Bar Association’s Joint Formal Opinion 2024-200 and the D.C. Bar’s Ethics Opinion 388, said Daniel J. Siegel, the principal of the Law Offices of Daniel J. Siegel and Integrated Technology Services.
“They all touch on the same issue,” said Siegel, who is also the chair-elect of the ABA Law Practice Division. “You have to read and check everything you do. You had to do it when you didn’t have computers.”
Siegel noted ABA Formal Opinion 512 provides similar guidance. He also said the ABA Model Rules of Professional Conduct and, in particular, Model Rule 1.1, which covers lawyers’ duty of competence, apply to lawyers’ use of generative AI.
Follow along with the ABA Journal’s coverage of the ABA Techshow 2025 here.
“At the end of the day, it’s competency,” Siegel said. “If you’re not reading your brief or you’re not reading your motion or you’re not reading your complaint and making sure everything is accurate, you’re not a competent lawyer.”
Siegel added that lawyers who use generative AI should also be aware of Model Rule 1.6, which focuses on the duty to protect client confidences.
According to Model Rule 1.6: “A lawyer shall make reasonable efforts to prevent the inadvertent or unauthorized disclosure of, or unauthorized access to, information relating to the representation of a client.”
Ivy Grey, the chief strategy and growth officer for editing software company WordRake, pointed out that information related to legal representation is often much broader than lawyers think.
“They think it’s only the stuff the client told them in private,” Grey said, pointing out that it can include metadata, sensitive facts and even algorithms. “But people aren’t thinking about these pieces of business information or personal information when they are using these publicly available generative AI tools.”
Grey added that password-protected information, including court records, should also be treated as confidential.
Before using generative AI tools, Wondracek suggested reading their terms of service to fully understand their privacy protection and data retention. She said this is especially important for ChatGPT and other general generative AI tools.
“If you are on a free version, don’t expect any privacy whatsoever,” Wondracek added.
See also:
At ABA Techshow, generative AI tools and how they help lawyers work faster took center stage
Write a letter to the editor, share a story tip or update, or report an error.
