Many surveyors love new tools and technologies, and artificial intelligence (AI) is one new technology that is being embraced by the profession. AI has been integrated into many facets of surveying, including hardware and software workflows1 of GNSS and sUAS.2 Surveyors may now also use AI to read deeds and produce legal descriptions3 and have even used it to research legal precedent.4
However, the use of AI to perform legal research poses real risks to surveyors who use it without safeguards in place. In this article, I first offer perspective on the risk of fabrications when AI is used to research the law or in the course of subject matter expert testimony. I then offer strategies for minimizing risk if AI is used in these ways.
Cautionary Tales: Attorneys and Expert Witnesses Both Fall Prey to AI Fabrications
One major concern about the use of AI in the context of legal research is that AI may fabricate information. In one instance, when I asked a popular large language model AI chatbot for a case summary, it fabricated the entire content of the case, from the nature of the dispute to the disposition of the case. I asked it to revise its answer twice more and it provided me with more fabricated answers. Eventually, it told me that it couldn’t summarize the case because it didn’t have access to legal research databases. In another instance, while the case summary was correct in general terms, it provided the incorrect location of the real property that was the subject of the lawsuit. A small but important detail when litigating boundary issues.
These AI fabrications are not isolated to my experience. As the use of AI has become more common, attorneys from jurisdictions throughout the country have been disqualified,5 fined,6 or sanctioned7 for using made-up citations or information generated by AI.
Perhaps more pertinent to many surveyors is that even subject matter experts retained for litigation can fall prey to AI fabrications. Experts’ reports and testimony have been struck or withdrawn for containing AI-fabricated information, such as false quotations,8 made-up journals,9 or fictitious citations.10
The overarching lesson from these cautionary tales is that AI may make up information to the detriment of the user. When this occurs in the context of litigation, the stakes can be high. As such, surveyors should exercise caution if using AI and employ safeguards to minimize the risks.
Strategies for Surveyors to Minimize the Risk Associated with the Use of AI
Surveyors desire to make defensible judgments based on good data and the most correct law. As such, you can employ the following strategies to offset the likelihood of perpetuating AI fabrications and minimize the risks if you choose to use AI to research the law.
Knowledge and education are key. Awareness that AI may fabricate information when performing research and analysis of the law better arms you to offset the hazards associated with AI fabrications. Discuss AI limitations and use cases with other surveyors and educate others often.
Verify any legal precedent obtained from AI. One method of verification is to obtain and read the legal opinion. There are many ways to find state and federal legal opinions, and probably the most easily accessible is Google Scholar’s case law search function. However, not all cases are available on Google Scholar and it may be hard to verify whether a case is still good law on the issue you are researching, as Google Scholar does not have a comprehensive citator service that tracks the history and subsequent decisions regarding the case. Databases specifically designed for legal research, such as Westlaw, LexisNexis, or Fastcase, have a larger repository of cases and more nuanced citator services.
Enlist an attorney. Legal research is complex and good law must be verified. Especially if you are involved with a particularly difficult or complex issue, the help of a competent attorney is well worth the price to avoid AI fabrications and any associated consequences to your reputation or licensure.
If you consult with an attorney, ask the attorney to provide acknowledgements or disclaimers. An acknowledgement of AI use should contain, at the least, details of how AI was used and whether information generated by AI was verified by a qualified human. If AI wasn’t used, the attorney may include a disclaimer stating that no artificial intelligence was used to create the document.
Consider using acknowledgements or disclaimers of AI use for your own work. Sources of information for the work you produce should be documented and verifiable. With the uptick of AI manipulated images, you may find it good practice to offer a suitable acknowledgement or disclaimer. This may provide assurance to clients, and the courts, of the integrity of the survey work.
Be mindful that inputting information into AI may create confidentiality or privacy concerns. Great care should be exercised in what you share,11 especially in the context of active litigation.
Depending on your state’s definition of the practice of law, exercise caution and confirm with legal counsel so that unauthorized practice of law does not occur. Consider seeking out an attorney to provide your professional organization with education on legal research, the definition of the practice of law, or insights regarding AI use for legal research or subject matter experts involved in litigation.
Conclusion
AI is not going anywhere. While AI appears to have many benefits, all that glitters is not gold. I hope the examples in this article serve as a deterrent for misuse of AI, rather than a warning of what is to come for the surveying profession as AI use increases. Exercise caution to safeguard your work, your reputation, and your clients.
You have your license, experience, and professional judgment, and you will likely be held liable for any AI-generated fabrications in your work. AI has no conscience, no morals, no ethics, and no emotions. AI won’t feel ashamed for providing you with false information, it won’t feel guilty for getting you in trouble, and it won’t be held accountable. Intelligence that is artificial, indeed.
Note: No generative artificial intelligence was used in writing this article. Information contained within this article is not legal advice and does not establish an attorney-client relationship. The opinions in this article are the author’s own and do not represent the position or opinions of the author’s employer.
From teaching university classes to digging fire line, Vailferree Brechtel has extensive education and experience in natural resources and environmental management, law, communication, conflict resolution, surveying, and literature. She is licensed in Alaska as an attorney and a Professional Land Surveyor. She works as a surveyor at the Alaska Department of Natural Resources, Division of Mining, Land & Water, where she tackles boundary issues at the intersection of surveying, law, and land management.
- 1 Top 5 AI Tools Transforming the Work of Surveyors
- 2 Modern Land Surveying: Revolutionizing Techniques with AI
- 3 Deed Reader Pro
- 4 AI at the Boundary – The American Surveyor
- 5 Judge disqualifies three Butler Snow attorneys from case over AI citations | Reuters
- 6 California issues historic fine over lawyer’s ChatGPT fabrications
- 7 Massachusetts Lawyer Sanctioned for AI-Generated Fictitious Case Citations | Maryland State Bar Association
- 8 NetChoice Pulls Expert After Allegations of AI Fabrications (1)
- 9 When AI Hallucinations Become Expert Testimony: A Warning Tale for Litigators | Maryland State Bar Association
- 10 Disqualifying Lawyers Over Expert’s AI Misuse Too Harsh, Plaintiff Says — Mealey’s
- 11 5 Things You Should Never Share with ChatGPT — AgileBlue