Advancing Computation Of Just Compensation: An AI-Based Approach To Property Valuation In Eminent Domain

by Molly McInnis, Associate Member, University of Cincinnati Law Review Vol. 94

I. Introduction

The United States Constitution’s Takings Clause guarantees that states will justly compensate a property owner whose property is taken by means of eminent domain.¹ This assurance protects individuals from the significant financial burden of losing their private property for the benefit of public use.² To determine the amount of compensation that is just, courts today use expert appraisers to value real property based on its fair market value.³ However, this practice is susceptible to error and risks the Constitution’s guarantee of just compensation.⁴

This article explores the potential benefits and risks of introducing artificial intelligence (“AI”) valuation methods to the process of eminent domain. Part II provides background on the importance of upholding the Fifth Amendment guarantee of just compensation and how the traditional method of property valuation falls short of doing so. Part III discusses the strengths and limitations of AI property valuation. That section also proposes solutions on how courts should approach the use of AI valuation to mitigate risk and maximize value. Finally, Part IV offers a brief conclusion, proposing an oversight approach by expert appraisers to AI valuation.

II. Background

Eminent domain is the power of the state to take privately owned property and convert it for public use.⁵ The Fifth Amendment of the United States Constitution limits the states’ power by requiring the government to provide the property owner “just compensation” for the condemned property.⁶ This guarantee is known as the Takings Clause, which ultimately protects individuals from bearing the financial burden of a publicly beneficial project.⁷ The idea of “just” compensation ensures that the property owner is compensated the “full and exact equivalent” of the property value they lose to condemnation.⁸ The Takings Clause serves as a vital tool in protecting the rights of property owners and limiting the government’s power.⁹ To preserve equity and fairness alongside the state’s taking power, it is critical for the justice system to ensure that property owners receive compensation that fully repays them for the property they have lost.

A. The Traditional Approach of Fair Market Value

Enforcement of just compensation protection is subject to judicial and legislative deference.¹⁰ The United States Supreme Court has routinely applied the fair market value standard when calculating just compensation in condemnation proceedings.¹¹ The fair market value approach intends to compensate the owner “what a willing buyer would pay in cash to a willing seller at the time of the taking.”¹² This procedure aims to situate property owners in the same objective position as if the property had never been taken.¹³

To determine the fair market value of real property, courts typically utilize expert appraisers, who issue appraisal reports and valuations of the property.¹⁴ Sometimes these experts are called into court to justify their calculations.¹⁵ The appraiser seeks to objectively value the property at which a willing and unpressured buyer would pay to a willing and unpressured seller in an “arm’s length transaction.”¹⁶ While the burden initially falls to the condemner (the state) to determine proper compensation, property owners may counter the state’s calculations by introducing their own evidence of fair market value.¹⁷ This process allows property owners to hire their own appraisers to advocate for what they think is fair compensation for their property.¹⁸

B. The Pitfalls of Expert Appraisals

Many situations complicate the computation of fair market value, making it increasingly difficult for appraisers to deliver just valuations.¹⁹ Given the inherent limitations of the current appraisal process, scholars argue that expert valuations can realistically only mitigate the extent of loss, rather than consistently produce perfect compensation for the deprivation of an individual’s private property.²⁰

Several factors lead to potential inaccuracies of the current appraisal process.²¹ In the absence of perfect data, the government is prone to undervaluing property.²² To calculate fair market value, an appraiser typically compares the value of the property being condemned to similar properties that have recently been sold on the market.²³ This process inherently requires appraisers to make assumptions regarding the value of the condemned property, risking the miscalculation of accurate valuations.²⁴ For example, if a home was passed down in a family for one hundred years, there may not be sufficient data available to reflect the property’s actual appreciation.²⁵ As a result, the calculated market value of that property tends to be speculative.²⁶ This speculation typically results in undervalued property when the home is condemned.²⁷ Additionally, inaccurate valuations are more likely to occur in a volatile market when the market values of homes rapidly fluctuate.²⁸ Consequently, datasets used for projecting market value may be incomplete or inaccurate amid turbulent real estate markets.²⁹ The limitations of appraisers may erode the accuracy and consistency of the traditional valuation process.³⁰ These concerns risk property owners receiving unjust compensation when their property is condemned.

Another issue identified within the traditional practice of valuation is the correlation between the expense and accuracy of appraisals. Studies suggest that correctness of home valuations may increase based on the amount of time and money expended for the appraisal.³¹ As a result, low-cost appraisals are more susceptible to error.³² Data suggests that the accuracy of property valuations improves when appraisers invest more time in selecting and analyzing comparable properties.³³ Since expert appraisers typically bill the by hour, the more time a client can afford directly affects the accuracy and precision of appraisal valuations.³⁴ When facing time and budget constraints, appraisers may be forced to rely on limited data, resulting in less accurate valuations. Additionally, more experienced appraisers charge higher rates for their degree of judgment and depth of expertise.³⁵ Producing accurate valuations relies on an appraiser’s ability to interpret data through reasoned logic and adherence to professional standards.³⁶ A more affordable, less experienced appraiser may lack the same depth of knowledge to produce comparably reliable valuations. Although property owners can challenge state valuations in condemnation proceedings, they may face financial or time limitations to hire a qualified expert to testify on their behalf.³⁷ Accordingly, a lack of resources may prevent property owners from countering state valuations that fall short of the true objective value of their property. This limitation creates another barrier to property owners in the process of obtaining just compensation.

C. The Benefits of an Artificial Intelligence Approach to Valuation

Today, many industries use artificial intelligence to value property in real estate sectors, such as insurance models and tax assessments.³⁸ While courts do not widely use AI-based valuation within condemnation proceedings, AI valuation models can overcome some of the shortfalls of traditional appraisal methods.³⁹ Ultimately, AI valuation models have the potential to improve the accuracy, efficiency, and affordability of property appraisals.⁴⁰

AI-based valuation models can produce more accurate property valuations than traditional appraisal methods. Scholars have analyzed the advantages of using AI property valuation for taxpayer liabilities.⁴¹ Similar to eminent domain, this approach to property valuation relies on the fair market value calculations.⁴² In contrast to limitations of traditional appraisals, AI models can sift through millions of variables while continuously updating data calculations.⁴³ This pattern creates increasingly accurate valuations based on changing market conditions over time.⁴⁴ AI valuation models are equipped to analyze all available information at the time of valuation, enabling them to detect complex relationships and nuanced patterns within large datasets without making implicit assumptions based on the property’s features.⁴⁵ Even expert appraisers are, by nature, prone to bias, error, and limited in the scope of data they can review.⁴⁶

AI models not only produce more accurate valuations than traditional appraisal methods, but they can do so in a more efficient and low-cost manner.⁴⁷ This capability can potentially lower the cost of accurate valuations by computing estimates in significantly less time. The sophistication of AI’s processing power may also reduce the variability of accurate AI valuations by aiding the work of less experienced and more affordable appraisers.

D. The Potential Risks of Artificial Intelligence Valuation

While the introduction of AI valuation methods may offer benefits in the eminent domain space, it is not without its downfalls. Left unchecked, AI models may fall short in providing transparency and consistency in property value computations. To reap the benefits of such programs, one must be aware of the inherent risk in AI valuation.

One issue with the AI-based valuation method is the lack of transparency and justifications of computations.⁴⁸ Traditionally, when an appraiser conducts a valuation, their credibility hinges on their compliance with the professional standards of conduct.⁴⁹ General compliance is set forth in the Uniform Standards of Professional Appraisal Practices.⁵⁰ These standard practices promote both transparency and credibility for property valuations.⁵¹ AI models, however, engage in complex computations and often produce results that are difficult to interpret, lacking the same degree of transparency.⁵² If the parties involved in the condemnation do not understand the methodology behind the valuation, its integrity is at stake.⁵³ Under the traditional appraisal approach, experts routinely justify their assessments through testimony of their rationale.⁵⁴ Conversely, using AI tools could potentially undermine the credibility of the appraisal process by creating dependence on a tool that even professional appraisers cannot explain.⁵⁵

Another possible downfall of AI valuation methods is the potential lack of consistency.⁵⁶ AI models produce their output valuations by detecting patterns within large amounts of data.⁵⁷ Lack of uniform reporting standards or inconsistent data compromises the credibility of AI valuation.⁵⁸ For instance, if AI models are fed non-uniform dimensions, measurements, or missing values, the output becomes inconsistent and unreliable.⁵⁹ This disparity could result in property owners receiving significantly different valuations for properties of comparable market values. Without consistent and standardized data, neither property owners nor appraisers could reasonably rely on these models to produce defensible valuations. Without such consistency, AI tools are just as prone to error as human appraisers.

III. Discussion

The unique capabilities of AI models offer increased accuracy, efficiency, and affordability. However, leaving such processes unchecked may jeopardize transparency and consistency within the condemnation process. To alleviate these concerns, utilizing expert oversight would allow the judicial system to both maximize the benefits and reduce the risk of AI valuation.

A. Promoting Just Compensation with Artificial Intelligence-Based Valuation within Eminent Domain

By introducing AI into the judicial process, property valuations may become significantly more accurate, better upholding the Fifth Amendment’s guarantee of just compensation. Given that AI models possess analytical capabilities that exceed the limitations of human appraisers, courts should consider incorporating AI-based valuation methods into condemnation. Research suggests the best approach is to integrate AI capabilities into the valuation process subject to strict oversight of expert appraisers.⁶⁰ AI should not replace the role of these experts but should rather become a tool in which they may apply their own expertise and supervision. By doing so, expert appraisers can analyze significantly larger datasets and identify nuanced patterns that human judgment alone might overlook. Because just compensation demands that valuations are both precise and consistent, incorporating AI into valuation calculations helps ensure these standards are met. More accurate AI-based valuations would also promote judicial efficiency by reducing the need for property owners to present counter valuations, which are often costly and time consuming. By utilizing AI’s sophisticated analytical capabilities to generate more consistent and accurate property valuations, courts better provide just compensation through enhanced accuracy and precision.

AI valuation methods also offer a more cost-effective approach than traditional appraisal valuations, benefiting both the condemning state and property owner. As noted, AI models can process and analyze significantly larger datasets more efficiently than human appraisers. By rapidly processing vast amounts of data, AI models can produce appraisals of equal or greater accuracy of human appraisers in significantly less time. This tool may also enhance the work of less experienced appraisers by increasing the accuracy and consistency of their less costly services. Such methods could be used by the courts when hiring appraisers to lessen the correlation between the cost and accuracy of valuations. Courts should adopt this cost-effective approach because it preserves financial resources while producing more accurate valuations.

AI valuation also benefits property owners by offering a more accessible resource to challenge state valuations in condemnation proceedings.⁶¹ For those unable to utilize expert appraisers, AI-based valuations may be a more affordable and accessible option. As a result, individuals may have the opportunity to advocate for just compensation when they would otherwise be limited in relying on potentially undervalued appraisals from the condemner. Ultimately, this approach strengthens the Fifth Amendment guarantee of just compensation by not only enhancing the accuracy of property valuations but doing so in a way that is more affordable for all parties involved.

B. Overcoming the Risk of Artificial Intelligence Valuation in Condemnation Proceedings

While the use of AI valuation promotes just compensation within the condemnation process, it also presents risks that must be managed to advance just compensation. The best approach to overcoming these AI-associated risks is ensuring strict expert oversight of AI incorporation.

As mentioned, the use of AI-based valuations could potentially undermine the transparency of property valuations. In the context of eminent domain—where objectivity and accuracy are vital to the process—it is fundamentally problematic for the government and property owners to rely on such a tool. If a property owner decides to dispute a government valuation in condemnation proceedings, they must hire their own appraiser to assert a reasoned and defensible valuation. If a property owner’s expert presents an AI-based valuation that lacks proper support, it cannot serve as a reliable basis for proposing the compensation owed. Consequently, the explanatory limitations of AI’s complex calculations may be a barrier to its ability to promote just compensation.

Developers have begun to address the issue of transparency or lack thereof.⁶² Studies have revealed the benefits of incorporating Explainable Artificial Intelligence (“XAI”) techniques into valuation practices. This AI model provides explanations of its decisions in certain outputs and valuations, citing factors such as zoning regulations, building age, and other components of computation.⁶³ When incorporating AI valuations into the condemnation process, courts should ensure that XAI methods are used. By including these explanations, XAI allows appraisers, hired by the state or property owners, to review valuation calculations and determine whether they comply with the professional standards and obligations of property appraisals.⁶⁴ This explanatory function not only benefits from the powerful computing capabilities of AI, but also requires the human oversight that ensures output values fall within a uniformly accepted procedure.⁶⁵ This approach promotes just compensation by delivering more accurate and consistent values backed by reason and defensible logic, mitigating the loss of transparency. In doing so, the use of AI valuations upholds the Fifth Amendment’s guarantee by advancing fairer and more accurate compensation within condemnation proceedings.

The issues of consistency in AI valuation outputs may also be remedied by expert oversight within the condemnation process.⁶⁶ As discussed, inconsistent data inputs undermine AI-based valuation’s ability to promote just compensation.⁶⁷ Inconsistent data could result in unpredictable valuations for property owners with comparable fair market value properties. The delivery of just compensation depends on valuations free from distorted and unreliable data. To mitigate this issue, courts should implement standardized reporting practices.⁶⁸ Such standardized procedures could be expanded to include the reporting of input data for AI valuation models.⁶⁹ This could be achieved through practices such as standardizing measurements and classifying variables.⁷⁰ When the state or property owner hires an appraiser for AI-based valuations, the appraiser should adhere to uniform standards. The judicial system should clearly define these standardized practices to ensure that expert appraisers can oversee the data input into the AI models. This process would support the consistency and credibility of AI valuation outputs while enabling experts to oversee the intricacies of the AI valuation process and facilitate reliable and precise data. This approach utilizes the benefits of AI and manages the potential risk of inconsistency.⁷¹ By maximizing the power of AI, property owners have a stronger chance of receiving more accurate and just compensation.⁷²

IV. Conclusion

The current practice of property valuation is not without its faults.⁷³ Introducing AI into the practice of condemnation offers substantial benefits such as affordability, efficiency, and accuracy.⁷⁴ Nonetheless, to maximize these benefits and avoid detrimental risks, the use of AI should be handled with careful expert oversight.⁷⁵ By ensuring proper oversight of AI Valuation processes, property valuation for eminent domain may more effectively achieve its goal in justly compensating individuals losing their property to condemnation. Courts should consider implementing these practices to uphold the Fifth Amendment guarantee that individuals affected by eminent domain are truly and fully compensated for their loss.

Cover Photo by Avi Waxman on Unsplash

Molly McInnis

References

1
U.S. Const. amend. V.
2
Monongahela Nav. Co. v. United States, 148 U.S. 312, 325 (1893).
3
Katrina Miriam Wyman, The Measure of Just Compensation, 41 U.C. Davis L. Rev. 239, 252 (2007).
4
Daniel B. Kelly, Acquiring Land through Eminent Domain: Justifications, Limitations, and Alternatives in Rsch. Handbook on the Econ. of Prop. L., 344, 353 (2022) (ebook).
5
Eminent Domain, Black’s Law Dictionary (12^th ed. 2024).
6
U.S. Const. amend. V.
7
Monongahela Nav. Co., 148 U.S. at 325.
8
Id. at 326.
9
Emilio R. Longoria, Properly Construing the Just Compensation Clause, 64 B.C. L. Rev. 1377, 1381 (2023).
10
James Geoffrey Durham, Efficient Just Compensation as a Limit on Eminent Domain, 69 Minn. L. Rev. 1277, 1278 (1985).
11
Wyman, supra note 3, at 252.
12
Id.
13
Id. at 253-54.
14
Dan L. Swango, Exploring Condemnation: Commentary on Real Property Valuation in Condemnation, 87 Appraisal J. 53, 54 (2019).
15
Id.
16
Fair Market Value, Corn. L. Sch.: Legal Info. Inst. (Feb. 2022), https://www.law.cornell.edu/wex/fair_market_value[https://perma.cc/W8K6-QATL%5D.
17
See United States v. Miller, 317 U.S. 369, 372 (1943).
18
Id.
19
Swango, supra note 14, at 56.
20
Alberto B. Lopez, Weighing and Reweighing Eminent Domain’s Political Philosophies Post-Kelo, 41 Wake Forest L. Rev. 237, 300 (2006).
21
Kelly, supra note 4.
22
Id. at 344.
23
Lopez, supra note 20, at 297.
24
Id.
25
Id.
26
Id.
27
Lopez, supra note 20, at 297.
28
Abdurrahman Yağmur Toprakli, AI-Driven Valuation: A New Era for Real Estate Appraisal, 18 J. of Eur. R.E. Rsch. 105, 105 (2024).
29
Id.
30
Id.
31
Kelly, supra note 4, at 353.
32
Id.
33
Alexei Alexandrov, Laurie Goodman & Michael Neal, Reengineering the Appraisal Process: Better Leveraging Both Automated Valuation Models and Manual Appraisals, The Urb. Inst. (Jan. 2023), https://www.urban.org/sites/default/files/2023-01/Reengineering%20the%20Appraisal%20Process.pdf [https://perma.cc/RWS6-89JJ].
34
Id.
35
Robert E. V. Kelley, Jr. & Hill Ward Henderson, Towards a More Credible Appraisal: An Attorney’s Perspective, Am. Soc’y of Appraisers (2025), https://www.appraisers.org/asa-newsroom/asa-blog/asa-blog/2025/03/21/towards-a-more-credible-appraisal–an-attorney-s-perspective-(arm-e-journal—2022—volume-6—issue-1)[https://perma.cc/8554-5L6K].
36
Id.
37
See United States v. Miller, 317 U.S. 369, 372 (1943).
38
Karthigeyan Kuppan, Deepak Bhaskar Acharya & Divya B., Foundational AI in Insurance and Real Estate: A Survey of Application, Challenges, and Future Directions, 12 IEEE Access 181282 (2024); Jay A. Soled & Kathleen DeLaney Thomas, AI, Taxation, and Valuation, 108 Iowa L. Rev. 651, 657 (2023).
39
4 Philip Nichols, Nichols on Eminent Domain §13.01 (2025).
40
Gabriela Droj, Anita Kwartnik-Pruc & Laurentiu Droj, A Comprehensive Overview Regarding the Impact of GIS on Property Valuation, 13 Intl. J. of Geo-Inf. (2024).
41
Soled & Thomas, supra note 38, at 685.
42
Id.
43
Id. at 689.
44
Id.
45
Juergen Deppner, Benedict von Ahlefeldt-Dehn, Eli Beracha, & Wolfgang Schaefers, Boosting the Accuracy of Commercial Real Estate Appraisals: An Interpretable Machine Learning Approach, J. of R.E. Fin. and Econ. (2023).
46
Soled & Thomas, supra note 38, at 689.
47
Id. at 686.
48
Chung Yim Yiu & Ka Shing Cheung, Enhancing Explainable AI Land Valuations Reporting for Consistency, Objectivity, and Transparency, 14(5) Land 297 (2025), https://doi.org/10.3390/land14050927, [https://perma.cc/BTQ5-2GLJ].
49
About the USPAP, Appraisers Assoc. of Am., https://www.appraisersassociation.org/education/uspap/about-uspap (last visited Sep. 26, 2025) [https://perma.cc/5GCE-WVK8].
50
Id.
51
Yim Yiu & Shing Cheung, supra note 48.
52
Id.
53
Id.
54
Id.
55
Id.
56
Id.
57
Droj, et.al., supra note 40.
58
Yim Yiu & Shing Cheung, supra note 48.
59
Id.
60
Id.
61
Miller, 317 U.S. at 372.
62
Id.
63
Id.
64
Id.
65
Id.
66
Id.
67
Id.
68
Id.
69
Yim Yiu & Shing Cheung, supra note 48.
70
Id.
71
Id.
72
Id.
73
Wyman, supra note 3, at 252.
74
Soled & Thomas, supra note 38, at 657.
75
Yim Yiu & Shing Cheung, supra note 48.