Generative Artificial Intelligence: Future Predictions Concerning Risk Management
By Sarah A. Johnson, Esq.
We have all heard predictions about how generative artificial intelligence (AI) is likely to change how professionals practice in both good and bad ways. Design professionals are not immune to such predictions. Indeed, engineers have already reported on the ability of AI to create plans and reports.
While we have not seen any legal cases specifically involving the use of generative AI in a professional liability lawsuit against a design professional, we anticipate that such cases will be forthcoming, especially as generative AI relevant to the engineering and architectural fields improves and becomes more prevalent. In general, there are many concerns involving the use of generative AI, including a lack of sufficient safeguards to ensure that the final product is based on sound engineering principles and accurate calculations.
Despite the risks involved with utilizing generative AI, it is likely that the field of engineering will eventually accept its use, just as it once did with other technological advances in the field that have now become commonplace, such as the use of Computer Assisted Drafting and Design (CADD).1 However, one thing that will almost certainly remain is that the engineer who signs and stamps a document will be responsible for its contents, whether or not generative AI was used in creating said document.2
Accordingly, it is important to approach generative AI with caution. While it may seem tempting to utilize generative AI to save time, any work product generated, in whole or in part, by generative AI should be reviewed and checked by the licensed professional, particularly when the AI is untested and/or the engineer is unaware of the sources utilized by the AI.
Further, we have seen an increase in “amateur” architects and engineers or unlicensed individuals generating plans, drawings, and reports over the years. These wannabe design professionals often seek out licensed professional architects and engineers to sign and stamp their generated documents.
As generative AI advances, it is anticipated that so will the requests for these types of services. This may seem like an easy way for an engineer to earn a small fee; however, in the age of AI, it will be more important than ever to view such requests with skepticism and either decline or, at the very least, thoroughly review and verify that any such document complies with sound engineering practices and is accurate and reliable. Engineers should also recognize that if the latter cannot be done in an economically viable manner, declination is the only option.
Apart from concerns about signing and sealing or otherwise utilizing documents created through the use of generative AI, several court cases have shown that battles over the use of copyrighted materials are another concern that comes with the use of the technology.3
Even though the law regarding copyright infringement involving generative AI is currently in flux, it would be prudent for architects and engineers to take extra care to protect their copyrighted materials from being used by generative AI. On the flip side, design professionals should be careful not to use copyrighted materials when utilizing generative AI, recognizing that it may not always be easy to tell whether an AI is doing so.
Although this article comments on some of the anticipated issues in the area of risk management that generative AI might cause, it is likely that many other unexpected issues may also surface, increasing the need for foresight when generative AI is utilized by architects, engineers, and design professionals.
1NSPE Board of Ethical Review Case 90-6, p. 3 (November 9, 1990) (finding that it ethical for an engineer sign and seal both documents he prepared using Computer Assisted Drafting and Design (CADD) and also documents prepared by another engineer using CADD working under the engineer’s direction and control).
2Id. (“The key requirement is that an engineer […] assumes the full responsibility for the work product.”).
3Am. Soc’y for Testing & Materials v. Upcodes, Inc., No. 24-2965, 2026 U.S. App. LEXIS 10137, at *31-32 (April 7, 2026) (finding that Up Codes, Inc. was likely to succeed in its defense of fair use with respect to its copying of ASTM Standards in its online research platform offering a searchable database of building codes); Thompson Reuters Enter. Ctr. GmbH v. Ross Intel. Inc., 765 F. Supp.3d 385, 401 (D. Del. 2025) (granting Thompson Reuters’ motion for partial summary judgment on copyright infringement of Westlaw headnotes by a competitor using the headnotes to train its AI).
