Should AI be Disclosed?

Recognize Ethical and Professional Standards of Using AI

AI is pervasive and promises to fundamentally transform business operations, education, and daily life. However, functional insights bridging the gap between AI and our industry have been lacking. The authors share their views on the ethical and professional standards implicated from the use of AI in forensics and valuation engagements.

This article explores the need for, or lack thereof, to disclose the use of artificial intelligence (AI) technology in the fields of business valuation and financial forensics. AI is pervasive and promises to fundamentally transform business operations, education, and daily life. However, functional insights bridging the gap between AI and our industry have been lacking.

During the December 2024 NACVA and the CTI’s Business Valuation and Financial Litigation Super Conference held in Las Vegas, we (Nick Mears and Nainesh Shah) aimed to bridge this gap by providing hands-on demonstrations for the “every day” analyst, manager, and owner. We stressed the importance of understanding the pros and cons of using AI technologies in practice. There has been much debate regarding whether the use of AI should be disclosed in work papers, schedules, or reports. As of today, the industry remains confused on this issue.

As we delve deeper into the nuances of AI, it is crucial to understand its fundamental aspects and their implications for our industry. The confusion around AI disclosure stems from the need to revisit and/or reinforce existing professional standards and ethical guidelines, ensuring they adequately address the use of AI technologies.

We encourage you to engage with this article, as it will highlight the potential of AI within our field while embracing the professional and ethical standards we must all adhere to. Therefore, this article will address the elephant in the room “should AI be disclosed?”

What is AI?

AI, broadly defined, includes technologies that perform tasks typically requiring human intelligence. Machine learning (ML), a subset of AI, focuses on training computer systems to recognize patterns in data, while deep learning delves further, enabling tasks like image and speech recognition, natural language processing, and prediction making. Generative AI, a more recent development, utilizes large language models (LLMs) to understand and generate text based on language patterns, creating new written, and can be integrated with other specialized models to create visual, and auditory content, exemplified by the popular ChatGPT.

How Can I Use AI in My Practice?

AI technologies in the valuation and forensic field include synthesizing complex insights, conducting comprehensive research, brainstorming ideas, and checking for calculation inaccuracies or misinformation. These technologies offer cost-effective solutions that expedite task completion and pinpoint logical weaknesses. AI can also be leveraged to reduce administrative tasks, streamline e-mail management, summarize PDFs, and meeting summaries. These tools can also help neutralize tones in communications, facilitate challenging client interactions, and generally improve efficiency. It is worth noting that the use of AI in administrative roles may come with certain drawbacks, such as the loss of a “personal touch.” However, these challenges are easily overcome, and their benefits far outweigh their shortcomings. A hands-on approach, like the one we did at the Super Conference, will be the right way to learn these tools.

What are the Limitations and Ethical Considerations?

A major concern when using AI technologies is the phenomenon of AI hallucinations, where the technology generates fabricated or incorrect information. AI technologies also may include biases, which are a symptom of their internal training data. Another significant limitation is the lack of common sense and real-world knowledge. In financial forensics, we call this knowledge, experience, education, training, and skills (KEETS). Lacking this knowledge means that AI technologies will struggle with basic concepts that humans find intuitive. AI also has difficulty with complex or multi-step reasoning tasks, which are often necessary for thorough business valuations. This might change in the future as the AI models improve.

As you can see, while AI is a powerful and useful tool, it is not without its problems. Practitioners that rely on these technologies without exercising oversight and professional judgment will likely result in skewed analysis and outcomes. This has been the focal point, which has been driving the “disclosure train.” However, acknowledging that these limitations exist will help you craft better AI-generated outcomes and ensure you are in control of the process.

Who is Responsible for Upholding Valuation Standards?

While concerns about verifying AI-generated content are valid, it is incorrect to blame all issues on AI. Failure to verify source documents or AI outputs violates professional standards across all valuation professional organizations (VPOs), regardless of AI usage. Qualified experts, even those who do not utilize AI, may deviate from professional standards yet still be permitted to testify. The underlying issue is a lack of accountability, whether stemming from the misuse of AI tools or biases in expert reports. In both scenarios, professional standards are breached, yet such experts are allowed to present their testimony. So, who should hold them accountable? For litigated matters, it should be the courts through voir dire and/or proper cross-examination. This can only be achieved if the trier of fact becomes aware of the expert’s professional and ethical standards. For non-litigated matters, it will be up to the regulatory bodies having similar knowledge.

Skepticism towards AI in valuations is valid, but experts must revisit the foundational professional and ethical standards for compliance. A competent valuation hinges on professional judgment, which should be the central test of whether services were performed adequately. This is crucial for admitting expert testimony, where opposing experts must clearly show how the lack of professional judgment impacted the valuation. Professional judgement is crucial for every VPO. The following table highlights how often “Professional Judgement” appears in the standards, indicating its importance. Simply put, a credible valuation cannot be completed without professional judgment.

What are the Industry Perspectives?

The NACVA announced the creation of the Artificial Intelligence and Machine Learning Commission (AIMLC). The mission is to synthesize the extensive AI knowledge base into clear, practical guidance, aiding in integrating AI innovations into valuation and litigation practices in a professional and ethical manner.[1] The AIMLC issued the first NACVA Advisory Brief, which was dedicated to the use of AI and ML.[2] The Advisory Brief worked to clarify the role of professional judgment when utilizing AI technologies. Professional judgment is inherent in our standards and was cited in the NACVA Advisory Brief, as well as other publications within our industry as shown below.

NACVA Professional Standards

A member/credentialed designee may rely upon information provided by any source without corroboration if disclosed in the report.[3]

Interpretation: The fact that a member may rely upon information provided by any source without corroboration if disclosed in the report has no bearing upon how a trier or arbiter of fact may rule regarding the reliability of an expert’s testimony. A trier or arbiter of fact may rule/find a member’s testimony or the underlying data (basis) of said testimony unreliable, less reliable, or lacking foundational reliability whether the member discloses their uncorroborated reliance upon data or not. If the underlying data of a member’s expert opinion is deemed unreliable, the expert opinion or conclusion drawn from said data may, by default, also be ruled, deemed, or found unreliable.[4]

As long as the source of the information is disclosed in the report, regardless of its source (i.e., Google search, publications, or an AI-generated response), the reliability is maintained. Interpreting this standard, it is evident that disclosing the use of AI is not necessary, just as disclosing the use of Google is not necessary. This is different than relying on the output without reviewing the source document or understanding the output.

NACVA Advisory Brief

Constituents should consider disclosing the use of any automated data output that was utilized as part of the analytical process and ensuring that such technologies were applied in an ethical manner while applying professional judgment and proper due diligence. The primary objective is to uphold the transparency of the valuation process rather than an exhaustive disclosure of all tools and processes involved.[5]

Interpretation: Utilizing automated data output, such as AI-generated forecasts, is acceptable if the expert reviews and verifies the source document. For instance, if AI forecasts a long-term nominal GDP growth rate of 4.33% based on the Livingston Survey, and the expert confirms this information, professional judgment is exercised, and professional standards are upheld. In such cases, the Livingston Survey should be disclosed, but no additional disclosure is necessary as the information has been verified.

Uniform Standards of Professional Appraisal Practice (USPAP) Q&A

Even when incorporating AI, developing credible assignment results requires the professional judgment of the appraiser.[6]

It is inconsistent to mandate AI disclosure when other technologies used in valuations, like software and databases, do not face the same requirement. Professional judgment, a cornerstone of valuation standards, should guide the use of all tools, including AI. Appraisers must apply this judgment when incorporating AI-generated content. Failure to do so deviates from the standards set by VPOs. Therefore, a clear understanding of professional judgment is crucial for the ethical and compliant integration of AI in valuations.

Be Responsible

While AI technologies present numerous advantages in synthesizing insights, conducting research, and enhancing operational efficiency, their limitations, and ethical considerations must be meticulously managed. The necessity of professional judgment, as mandated by valuation standards, cannot be overstated in the integration of AI into the valuation process. By maintaining awareness of AI’s potential biases and inaccuracies, practitioners can effectively mitigate risks and ensure accurate, reliable outcomes. Ultimately, the responsible and informed use of AI will enable valuation professionals to harness its benefits while upholding the highest standards of integrity and expertise in their practice.

[1] NACVA Association News, NACVA Announces New Artificial Intelligence Commission – First Quarter 2024.

[2] NACVA Advisory Brief: The Use of Artificial Intelligence and Machine Learning.

[3] NACVA Professional Standards Section IV.C Reliability of Data.

[4] NACVA Standards and Ethics FAQ Library, Litigation Question 2, Reliability of Data.

[5] NACVA Advisory Brief: The Use of Artificial Intelligence and Machine Learning.

[6] The Appraisal Foundation, 2024 USPAP Q&A, 2024-02 Artificial Intelligence (AI), issued January 16, 2024.

---

Nick Mears, MBA, CVA, MAFF, is the founder and managing member of Caprock Business Consulting, LLC, a business valuation and forensic litigation services practice. He has completed over 1,000 detailed business valuation reports over the past 13 years and specializes in business valuation and financial forensics in the areas of family law, litigation, and shareholder disputes, transaction consulting, SBA and commercial lending, and gift and estate tax matters.

Mr. Mears is currently a faculty instructor for the National Training and Development Team for the NACVA. He was the previous Chair of the AIMLC in 2024, is a current member of the Ethics Oversight Board (EOB), and was previously a Standards Board (SDB) member.

Mr. Mears can be contacted at (806) 853-7832 or by e-mail to nick@caprockbc.com.

Nainesh Shah, CFA, CVA, is a seasoned professional with over 25 years of experience in business valuation, investment management, and financial services. He is the co-founder and director of Valuation Advisory at Complete Advisors, where he specializes in the valuation of complex and intangible assets. His expertise has led him to design and execute sophisticated valuation models for a wide range of assets, including public and private companies, patents, and private equity.

Beyond his professional accomplishments, Mr. Shah actively contributes to the field as a member of the NACVA Standards Board (SDB) and AIMLC. He also holds the CFA designation and is committed to social impact through his involvement with Upaya Social Ventures.

Mr. Shah can be contacted at (516) 240-6162 or by e-mail to nainesh@completeadvisors.com