Laura B. Angelini, Damien C. Powell, and Julianna Malogolowkin share recent insights with Law360 on the need for attorneys to understand the changing AI legal landscape.
The artificial intelligence revolution is well underway, and recent litigation highlights the need for attorneys to understand the evolving AI legal landscape in order to provide sound guidance to clients across all industries. A recent lawsuit stemming from the use of AI in an academic setting is instructive.
In late September, parents of a high school student in Hingham, Massachusetts, sued the Town of Hingham School Committee, teachers and administrators — in Harris v. Adams, pending in the U.S. District Court for the District of Massachusetts — after their son, who is currently a rising senior at Hingham High School, was disciplined and received a D on a project due to his use of AI. The minor son is referred to as “RNH” in court filings.[1]
The plaintiffs sought various forms of injunctive relief to prevent irreparable harm to RNH’s college admissions prospects, future career opportunities and earning potential.
Using AI to Prepare for a Project
In December 2023, while RNH was a junior, he and another classmate teamed up for an Advanced Placement U.S. history project.
According to the lawsuit, the teacher who assigned the project did not explicitly prohibit the use of AI to prepare for or research the project. RNH and his classmate used AI — including an online AI application, Grammarly — to generate content for research notes and a draft documentary script for the project that they submitted to their teacher as part of the assignment.
While the submission did contain citations, it did not cite any AI tools. When the students submitted the script — one component of the six-component assignment — via Turnitin.com, a website that contains a plagiarism detection tool, the website flagged parts of the submission as AI-generated. After following up on Turnitin.com’s findings, the teacher accused the students of cheating on the assignment due to their use of AI for preparation and research.
The teacher required RNH to redo the assignment on his own, without using AI or any of his previous work. However, due to his prior use of AI on the research notes and draft script, the teacher awarded him zero points for those components, which factored into a final grade of a D. The school also required RNH to attend Saturday detention and allegedly barred his admission to the National Honors Society.
No Written AI Policy
The merits of the lawsuit hinge on the absence of any AI-specific policy in place at Hingham High School at the time of the alleged misconduct.
While the high school’s student handbook generally prohibited academic dishonesty, cheating and plagiarism on school assignments, including a prohibition on the “unauthorized use of technology during an assessment,” it was entirely silent and contained no outright prohibition on the use of AI.
Moreover, none of the written instructions accompanying the assignment mentioned anything about AI or expressly prohibited its use.
The outcome of the litigation likely will turn on the questions of whether the school properly warned students that the use of AI tools in this context was prohibited and, relatedly, whether using AI in this context constituted academic dishonesty, as defined under the version of the student handbook in effect at the time of the alleged misconduct. The Hingham School Committee has since revised the academic integrity provisions of the student handbook expressly to address the use of AI.
Civil Rights at Stake
The lawsuit alleges, among other things, that the defendants violated RNH’s civil rights through its disciplinary actions, and that those actions have harmed and will continue to harm RNH’s college prospects, potential future earnings, earning capacity, income and career potential.
The plaintiffs petitioned the court for a preliminary injunction that, if granted, would have required the defendants to raise RNH’s grade in AP U.S. history from a C-plus to a B — eliminating the demerits he received for using AI — and expunge any disciplinary record. The plaintiffs also sought an order from the court that would retroactively induct RNH into the National Honor Society and require the defendants to undergo training for AI use in the classroom.
The plaintiffs argued that RNH would suffer permanent, irreparable harm if the court did not grant an injunction prior to a final decision on the merits of the lawsuit, given the fact that RNH is applying early decision to a number of elite colleges and universities, and the college admissions process is already well underway for the year.
An Uphill Battle
U.S. Magistrate Judge Paul G. Levenson held a preliminary injunction hearing on Oct. 22, 2004, in Boston. Each side presented arguments and witnesses during the nearly daylong hearing.
The plaintiffs argued that a school’s policies — here a student handbook that did not specifically reference AI — serve “as a vital tool in establishing clear expectations for behavior, conduct, and discipline among students.” The rules, policies and guidelines therein provide “a consistent framework that helps maintain order and fairness” within a community.
Because there was no clear policy prohibiting the use of AI, the plaintiffs argued that the school imposed discipline that was arbitrary, capricious and an abuse of discretion.[2] RNH testified to these points during the hearing. The counsel for the plaintiffs further argued that the use of writing generated by an AI tool does not constitute plagiarism because it is not another person’s work.[3]
The defendants countered by pointing to several different school policies and presentations that, they argued, did put students on notice about the restrictions on the use of AI.[4] Further, the defendants argued that RNH’s actions clearly constituted cheating and plagiarism, which the academic integrity provision of the student handbook prohibits.[5]
Finally, the defendants argued that Massachusetts Supreme Judicial Court precedent holds that courts should give substantial deference to schools in disciplinary matters.[6] During the hearing, Judge Levenson noted the lack of case law support for the plaintiffs’ position, and signaled that the plaintiffs had an “uphill battle” to succeed.[7]
Plaintiffs Not Likely to Succeed on the Merits of Their Case
On Nov. 20, the court denied the plaintiffs’ motion for a preliminary injunction in a 47-page decision.[8] In weighing the relevant factors, the court determined that the plaintiffs were not likely to succeed on the merits of their claims.[9]
The court found that there was nothing in the preliminary factual record to suggest that Hingham High School officials were hasty in concluding that RNH had cheated, in violation of the school’s academic integrity rules, and that any student in RNH’s position would have understood as much.[10]
Nor, the court found, were the consequences the defendants imposed so heavy-handed as to exceed the defendants’ considerable discretion in such matters.[11] The court also found it “doubtful” that it had any role in second-guessing the judgments of teachers and school officials who are responsible for grading and disciplinary decisions, particularly discipline short of suspension, and that RNH and his parents were afforded prompt notice of the school’s findings and were given the opportunity to be heard.[12]
While the court acknowledged there was “considerable force” to the plaintiffs’ irreparable harm arguments, in denying the preliminary injunction, the court found that public interest and a balancing of the equities, in addition to the unlikelihood of the plaintiffs succeeding on the merits, weighed in favor of denying injunctive relief.[13]
AI Is Everywhere
Regardless of the final resolution, the lawsuit highlights the reality that academic institutions and businesses must address head-on the proliferation of AI and the legal risks that accompany it.
It also underscores the need for attorneys to educate themselves on generative AI technology and the specific AI tools and applications that affect their clients’ businesses and/or institutions, so they can help clients establish clear expectations and limits around the permissible use of AI and assist in drafting transparent AI policies.
Education Is Key
Counsel can only help clients guard against the risk of litigation if they understand the uses of AI in the clients’ specific industries. The first step is for counsel to educate themselves about the various AI applications available to their clients. Only then can counsel help clients evaluate the associated risks and rewards and determine how to specifically address AI use in their specific environments.
Key considerations include whether the client should ban AI use completely or, instead, set clear guardrails around the types of AI that can be used, when it can be used, who can use it, the type of data that can be input or generated, authorization needed for use and the proper review process for AI-generated work.
For example, many companies have completely prohibited the use of generative AI tools or established clear limits on the specific tools and types of information that can be entered into these tools due to concerns about the risk of exposure of confidential and proprietary information and/or trade secrets. In the academic setting, a total ban on AI use is likely inappropriate for a host of reasons, including the fact that AI is already used day-to-day by students, whose future employers likely will expect them to be experienced in the use of AI.
A total ban also would ignore the reality that certain AI tools, when used appropriately, add positive value in the academic environment and support student learning.
Implement and Integrate Clear AI Policies
Once a client determines its expectations and direction around AI, attorneys should assist the client in creating clear and comprehensive AI guidance and incorporate that guidance, where appropriate, into the client’s existing compliance framework, e.g., academic integrity, codes of conduct and business ethics, human resources and discipline policies, security and data protection, data retention and data disposal.
Attorneys also should assist in creating educational and training materials that will elucidate the client’s work force as to expectations surrounding AI, including permissible and prohibited uses, and provide practical examples.
Stay Informed and Up-to-Date
The Hingham High School Lawsuit is a harbinger of what the legal community can expect to see in the evolving world of AI.
Like the parents in the lawsuit, plaintiffs in employment litigation and professionals in disciplinary cases will undoubtedly be asserting the same argument that the absence of clear AI guidance or ambiguity in employee handbooks or codes of conduct fail to establish clear expectations for behavior, conduct and discipline.
Moreover, Judge Levinson’s comments about the dearth of case law in this field emphasizes the need for counsel to stay up to date on cutting edge developments in the quickly evolving AI legal landscape. Implementing and enforcing AI policies is a good start, but clients and their counsel should be vigilant and willing to nimbly keep pace with new developments and standards in the AI community in order to best minimize the risk of litigation.
[1] Dale and Jennifer Harris, as Parents of RNH v. Margaret Adams et al., Civil Action No. 1:24-cv-12437-WGY.
[2] Jane Rosenweig, The First Case of Many, Slate (October 26, 2024), https://slate.com/life/2024/10/ai-chat-gpt-students-cheating-detection.html; Plaintiffs’ Supplement Memo of Law in Support of Motion for Preliminary Injunction (Dkt. 23)
[3] Id.
[4] Id.
[5] Defendants’ Opposition to Plaintiffs’ Motion for a Preliminary Injunction (Dkt. 24).
[6] Defendants’ Memorandum of Law in Support of Defendants’ Motion to Dismiss Plaintiffs’ Complaint Pursuant to Fed.R.Civ.Pr. 12(B)(6) (Dkt. 13).
[7] Brian Dowling, Uphill Battle To Fix Grade Docked For AI Use, Judge Says, Law360 (October 22, 2024), https://www.law360.com/articles/1925516/-uphill-battle-to-fix-grade-docked-for-ai-use-judge-says.
[8] Memorandum and Order on Plaintiffs’ Motion for Preliminary Injunction (Dkt. 30).
[9] Id.
[10] Id.
[11] Id.
[12] Id.
[13] Id.
