The AI Revolution is Coming for Business Valuation and Forensic Accounting

How AI is Transforming the Industry

Artificial Intelligence (AI) represents a paradigm shift in the realm of computational capabilities, transcending traditional boundaries of machine functionality to emulate human-like intelligence. AI has made remarkable advances in recent years due to the availability of massive amounts of data, powerful computing resources, and innovations in methods and architectures. This article will explore the role and impact of AI in valuation and forensic accounting, addressing the challenges and issues with AI, and ensuring quality control in using AI.

Introduction to AI in Business Valuation and Forensic Accounting

Artificial Intelligence (AI) represents a paradigm shift in the realm of computational capabilities; transcending traditional boundaries of machine functionality to emulate human-like intelligence. AI has made remarkable advances in recent years due to the availability of massive amounts of data, powerful computing resources, and innovations in methods and architectures.

One of the most exciting types of AI is Generative AI, which can create new data that is similar to a given dataset, such as images, text, or music. According to a McKinsey & Company study, Generative AI could unlock the potential annual value of $2.6 trillion to $4.4 trillion for 63 use cases across 16 business functions, ranging from personalized marketing e-mails to realistic video game environments. AI can also partner with valuation and forensic accounting professionals to enhance their efficiency, accuracy, and insights. By automating data collection and processing, generating calculations and reports, and improving decision-making and communication, AI can offer valuable support. However, it is essential to have human supervision to ensure the quality and reliability of the outcomes.

This article will explore the role and impact of AI in valuation and forensic accounting, by:

1. Analyzing its effects on business value;
2. Explaining its functionality and how it works;
3. Describing prompt engineering;
4. Addressing the challenges and issues with AI; and,
5. Ensuring quality control in using AI.

This article was created by:

1. Uploading presentation slides and notes from the author’s presentation at the 2023 NACVA and the CTI’s Business Valuation & Financial Litigation Super Conference to ChatGPT, which then created an outline for the article.
2. After adjusting the outline, the author prompted ChatGPT to write the sections of the article.
3. The resulting article was then copied and pasted into Microsoft Word, and then the author worked with Microsoft Copilot in Word and ChatGPT to rewrite each section.
4. The entire article was then edited, proofread, and fact checked by ChatGPT and traditional human-based techniques.

AI’s Impact on Business Value

In the ever-evolving landscape of business, AI stands as a beacon of transformation; influencing not just one, but multiple facets of business performance, strategy, and competitiveness. Let’s dive into the ways AI is reshaping the business world.

1. AI as a Technological Game-Changer: Think of AI as the master key unlocking the potential of other technologies. It turbocharges the development and innovation in fields like cloud computing, the Internet of Things (IoT), blockchain, and biotechnology. By offering groundbreaking tools for data analysis, optimization, and synthesis, AI is not just another cog in the wheel; it is the engine driving technological evolution.
2. Unleashing Economic and Productivity Powerhouses: The story of generative AI is one of hope and revolution. In a world grappling with economic slowdowns, marked by the 2012 stagnation and exacerbated by challenges like the COVID-19 pandemic and demographic shifts, AI emerges as a rescuer. Picture this: in banking alone, AI could add a staggering $200 billion to $340 billion in annual value (i.e., productivity gains, revenue enhancements, etc.). In retail? We are looking at an increase of $400 billion to $660 billion in value. Beyond numbers, AI is a time liberator. Tasks taking up a quarter of our work hours could be automated, making 60% to70% of work activities more efficient. It is not just about doing things faster; it is about transforming how we work.
3. Transforming Business Functions with Precision: The impact of AI is not a broad stroke; it is precise and function-specific. In customer operations, marketing, sales, and even software engineering, AI is the secret ingredient for enhancing outcomes. Sales productivity could see a boost of 3% to 5% in terms of annual global expenditures. In software engineering, productivity might leap by 20% to 45%. Imagine software developers using AI tools to complete tasks 56% faster. In research and development, AI could lead to cost savings of 10% to 15%. It is not just about doing things better; it is about redefining the possible.

The story of AI’s impact on business is not uniform. It varies dramatically across industries, sectors, and markets. Some, like media, telecommunications, and financial services, are ripe for AI’s transformative touch. Others, such as healthcare, education, and government, face more hurdles. The extent and nature of AI’s influence also hinge on the unique business model, strategy, and culture of each organization; not to mention external factors like regulations, competition, and customer preferences.

AI’s Functionality: How Does AI Work?

AI, particularly in the realm of machine learning, is like teaching a computer to understand and learn from data on its own, without needing step-by-step instructions. A key part of this is deep learning, which involves a complex structure known as a neural network.

Neural networks are inspired by the human brain and are made up of layers of nodes, or “neurons.” Each layer receives input, performs simple calculations, and passes its results to the next layer. The network processes information through these layers, learning to recognize patterns and make decisions.

One fascinating area of deep learning is creating large language models (LLMs) like GPT-4. These models are trained on vast amounts of text data and are structured to understand and generate human-like text. They analyze and learn the relationships between words in a sentence or across sentences.

When LLMs process text, they convert words into numerical representations, often referred to as vectors, in a high-dimensional space. This “vector space” allows the AI to quantify and analyze the relationships of words, such as the context or meaning of words and phrases. Each word or phrase is represented as a point in this space, and the spatial relationships between these points help the model understand language semantics and syntax.

Humans live and perceive in a world of four dimensions—three of space (length, width, height) and one of time. In our daily lives, we understand and navigate these dimensions quite intuitively. For instance, when a person walks from a house to the store, he or she moves through three-dimensional space over a period of time.

Now, consider GPT-4. Instead of operating in our familiar four-dimensional space, it works in a space with potentially hundreds or even thousands of dimensions. Each dimension can be thought of as representing some aspect of linguistic information; such as the meaning of words, their grammatical roles, the tone, the context they are used in, and so on.

An example to illustrate this could be the word “bank.” In our three-dimensional world, we understand “bank” based on context; it could mean the edge of a river or a financial institution. GPT-4, in its high-dimensional space, represents “bank” as a vector. The exact position of this vector in the high-dimensional space is influenced by all the other words and contexts in which “bank” appears throughout the training data.

In practical terms, when you provide a word, phrase, or prompt to an LLM, it maps this input into its vector space, processes it through its neural network layers, and generates an output. This output is then translated back from numerical form into human-readable text. The AI can perform a variety of tasks, such as writing, answering questions, or translating languages, based on the patterns it has learned.

The sophistication of models like GPT-4 lies in their ability to process and generate coherent, contextually relevant text, making them versatile tools for a wide range of language-related applications.

Prompt Engineering: Crafting Effective AI Queries

Prompt engineering involves crafting specific and effective questions or prompts to get the best responses from AI systems. This skill is crucial in AI usage, requiring a balance between detail and openness to leverage the AI’s analytical and creative abilities. It involves using precise, context-sensitive language to guide the AI towards the desired result.

This technique is a collaboration between human input and AI output. Improvement in prompt engineering can be achieved through using AI itself, like LLMs for generating or evaluating prompts, or tools like Microsoft Copilot for coding and writing assistance. Best practices in prompt engineering include using examples, step-by-step guidance, comparative analysis, and industry-specific inquiries.

In fields like business valuation and forensic accounting, prompt engineering enables applications like generating reports, data analysis, expert advice, and innovation. It also aids in professional learning by allowing interaction with AI in a natural way.

Examples of well-crafted prompts in these areas could be:

For Business Valuation:

“Generate a detailed report comparing the financial health of Microsoft and IBM over the last five years utilizing Yahoo! Finance, focusing on key metrics such as revenue growth, profit margins, and market share. Provide authoritative citations for the data used in the analysis.”

“What are the potential impacts of recent market trends on the valuation of a technology startup in the renewable energy sector? Provide authoritative citations for the market trends.”

For Forensic Accounting:

“Identify transactions in the uploaded checking account statements that exceed $1,000 (‘Large Transactions’), use the descriptions for each Large Transaction in the checking account statements to search the internet, and then create a new Excel file with the date, description, amount, and internet results for each Large Transaction.”

“Identify all transactions in the company’s uploaded general ledger that appear to be for meals and entertainment and enter the amount, date, and description for each of these transactions into a new Excel spreadsheet.” There are also examples on YouTube of how to analyze a general ledger with ChatGPT.

These prompts are designed to be specific, context-aware, and tailored to elicit detailed and relevant responses from AI.

Addressing the Challenges and Issues with AI

AI is not a perfect or flawless technology, and it comes with various challenges and issues that need to be addressed and resolved to ensure its ethical, responsible, and effective use in valuation and forensic accounting. Some of the challenges and issues include:

• Hallucinations: AI systems, especially LLMs, can sometimes generate text that is not based on facts or reality, but rather on imagination or speculation, which can lead to false or misleading information, such as inventing numbers, data points, or events, that are not supported by evidence or logic.
• Bias and fairness: AI systems can sometimes reflect or amplify the bias and unfairness that exist in the data or the society, which can lead to discrimination or harm, such as favoring or disfavoring certain groups, individuals, or outcomes, based on irrelevant or inappropriate factors, such as gender, race, or age.
• Lack of understanding: AI systems can sometimes lack the deep and comprehensive understanding of the data or the domain, which can lead to errors or inconsistencies, such as misinterpreting or ignoring the context, nuance, or intent of the data or the task, or failing to capture the complexity or diversity of the data or the domain.
• Data privacy: AI systems can sometimes compromise or violate the data privacy of the users or the clients, which can lead to exposure or misuse, such as collecting, storing, or sharing the data without consent or authorization, or using the data for purposes other than the intended or agreed ones.
• Intellectual property claims: AI systems can sometimes raise or challenge the intellectual property claims of the creators or the users, which can lead to disputes or conflicts, such as determining the ownership, rights, and responsibilities of the data, the code, or the output, that are generated or used by AI systems.
• Potential misuse: AI systems can sometimes be used for malicious or harmful purposes, which can lead to damage or danger, such as manipulating, deceiving, or attacking the users, the clients, or the society, by using AI to generate fake or fraudulent data, reports, or images.

These challenges and issues require careful and constant attention and action, from the developers, the users, and the regulators of AI systems, in order to ensure the trustworthiness, accountability, and transparency of AI in valuation and forensic accounting.

Ensuring Quality Control in AI Utilization

As professionals in valuation and forensic accounting, we have the opportunity to use AI to enhance our work and deliver better results. AI can help us analyze large and complex data sets, automate routine tasks, and generate insights and recommendations. However, to use AI effectively, we need to communicate clearly with our clients and stakeholders, manage the risks and limitations of AI, and follow the best practices and standards of our fields.

We also need to be aware of the challenges and issues that AI poses, such as ethical, legal, and social implications, and how to address them. AI is not a replacement, but a partner, for us to leverage and innovate. We need to adhere to our professional standards and ensure that our use of AI is transparent, reliable, and consistent.

Conclusion: The Future of AI in Valuation and Forensic Accounting

As AI develops and advances, it will bring a radical change to business valuation, forensic accounting, and business in general. We will likely be using AI in our daily valuation and forensic accounting work in the coming years. However, AI also poses various challenges and issues that need to be addressed and resolved, such as hallucinations, bias, lack of understanding, data privacy, intellectual property claims, and potential misuse. Therefore, it is important to ensure quality control in AI utilization. The future of AI in professional practice is not a replacement or a threat, but rather a collaboration and an opportunity, for valuation and forensic accounting professionals, to leverage the power and potential of AI, and to advance and innovate in our fields.

---

Scott DeMarco, CBA, CVA, CDFA, CPVA, is the owner and founder of Equitable Value LLC and has been qualified and appointed by courts as an expert witness, and his business valuation opinions have been adopted by courts. Prior to his career in business appraisal, he accumulated experience in finance, accounting, and marketing through his involvement as CFO in the growth and subsequent sale of two software enterprises. He is a Certified Business Appraiser, Certified Valuation Analyst, Certified Divorce Financial Analyst, and Certified Patent Valuation Analyst.

Mr. DeMarco can be contacted at (800) 601-0635 or by e-mail to scott@equitable.expert.