AI Hesitancy-Look Before You Leap
The accounting profession has always evolved. The profession has always been cautious, and AI hesitancy makes you look before you leap. However, the pace of change now is enabled by practical artificial intelligence (AI). Now, and in the future, we will utilize Large Language Model (LLM) general AI (GenAI), various products that include AI, and Agentic AI. As buyers, we will use specific products for our primary tasks, which are AI-enabled or leveraged, and GenAI for productivity and general tasks. Further, we will not consider AI as much once it becomes common. However, in the short term, many products without AI capabilities claim to have them. Management has asked the product managers, marketing teams, and sales teams to promote non-existent AI. The new term, AI-washing, is better than my old term, Fake AI. The debate about what AI is and what it is not should be straightforward, but a lack of understanding of AI algorithms and technologies makes it easy for misrepresentations to occur. As accounting professionals, you may not have a deep understanding of the computer jargon and technologies being discussed. We will provide background information to help you understand the opportunities, the need, and how to deliver the services effectively and efficiently.
Is The AI You Are Considering Marketing, Or the Real Deal?
To start, let’s provide a few definitions. Computer science defines AI research as the study of “intelligent agents,” any device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals. Colloquially, the term “artificial intelligence” is used to describe machines that mimic “cognitive” functions that humans associate with other human minds, such as “learning” and “problem-solving.” Much of today’s AI technology comes from deep learning algorithms and neural networks in the 2010 timeframe. On the other hand, Machine learning (ML) is a subset of artificial intelligence. ML is focused on enabling programs to learn from past data without requiring explicit programming. In short, ML applies historical information when processing today’s data without explicit programming. Almost all practical and useful examples of AI today are machine learning. While ML might seem complex, many tools and applications use ML every day. The main layers of Artificial Intelligence are: 1) Artificial intelligence is the overall technology category and the most inclusive term. 2) Machine Learning allows a computer to learn from past data without explicitly programming. 3) Neural networks mimic the human brain through a set of algorithms. 4) Deep learning is a neural network with more than three layers, with more automation.
Consider How Your Firm Should Deploy AI To Accomplish These Items
1) Start by creating a firm AI policy.
2) What LLM/GenAI platform should the firm use (Copilot 365, ChatGPT, Claude 3, Gemini, and others)?
a. How do I keep client data safe? This goes beyond simple items like not including any Personally Identifiable information (PII)
b. It may mean more extensive data protection, like Microsoft’s Enterprise Data Protection in Copilot 365
i. Authentication: It uses Microsoft Entra ID to ensure only authorized users can access Copilot with commercial data protection.
ii. Data Encryption: Chat data is encrypted both in transit and at rest using advanced encryption standards.
iii. Privacy: Prompts and responses are not saved or used to train the underlying models//Microsoft has no “eyes-on” access to this data.
iv. Anonymity: Search queries triggered by prompts are not linked to users or organizations.
v. Ad Protection: Advertising shown to users is not targeted based on their work identity or chat history.
3) Can you name things firms should and should not do with AI?
a. Use AI to search instead of Google, DuckDuckGo, or other search engines?
b. How can you reduce the learning curve for my team (prompt engineering)?
c. Have we standardized an AI prompting approach to leverage GenAI products?
d. Configure and customize AI for your preferences and common tasks to make the results more applicable to my situation?
e. How does integration into my typical applications work?
4) What are practical things I can do?
a. Filter incoming emails to prioritize client inquiries and urgent matters, helping administrative staff and professionals respond more efficiently.
b. Prepare template responses with generic examples for common client questions, enabling your professional team to streamline communication and maintain consistency.
c. Create customized marketing and social media posts from firm-specific content, allowing marketing teams to boost engagement and promote services effectively.
d. Develop slide decks and handouts tailored to client meetings, equipping presenters and advisors with professional materials that address each client’s unique needs.
e. Are Large Language Model (LLM) Generative AI products sufficient, or should we look for AI that has been implemented in the products?
5) Now that I understand the fundamentals, how do I accomplish more advanced tasks?
a. Pick areas of the firm with the most potential for improvement.
b. Ask existing vendors about any product improvements/AI plans.
c. Consider new competitors in each category of need.
d. Vet the license/privacy/sub-processor relationships for data privacy.
e. What AI governance should be applied?
Why AI, Why Now, And How?
1) Why AI – properly controlled AI can leverage general knowledge and specific opportunities for the client. Furthermore, AI assistants will quickly and comprehensively identify, gather, and organize client opportunities in ways that were previously unaffordable or impossible.
a. In private platforms, client confidentiality is protected, unlike many Large Language Model (LLM/Generative AI) products.
b. Known and effective business cases are structured to support legal approaches.
c. Accuracy can be controlled, minimizing hallucinations common in LLM products.
d. The time to construct and analyze solutions is reduced. Accounting professionals can utilize their skills to identify and prioritize solutions for clients, rather than selecting from a few approaches that have been reformatted for another client.
e. AI inside a platform has already been configured to optimize your professional work.
2) What do clients want?
a. Specific advice for their evolving needs. The needs of each client vary widely, and you need a methodology to capture and understand these needs. Clients’ needs progress over time, and your advisory solutions should also evolve accordingly. While today might involve operational excellence and minimizing tax exposure, children’s education, business succession, charitable giving, and preserving long-term wealth could become new needs.
b. Meeting specific needs, which could involve goals beyond compliance and profitability.
c. Maximizing profit
d. Minimizing risk
3) Why a platform rather than my own models?
a.
While we respect the thoroughness, creativity,
and inventiveness of accounting professionals, it is hard to leverage the
skills of relatively few people across your firm and all the clients using LLMs.
Further, transferring the specialized knowledge is challenging. Finally, there
may be many other opportunities that are not being presented to the client.
b.
Proprietary models are often constrained by
spreadsheet and word processing constructs.
c.
Financial information may have to be re-keyed
with changes in financial position affecting the strategy. The best models are
never static.
Your Role, Should You Decide to Accept It
You must protect client data, but your learning should already be underway. We have observed a slow adoption by accountants of various technologies over the years. We understand the bleeding edge, leading edge, early adopters, late adopters, and laggards. We remind you that the most competitive advantage comes from being a leading-edge adopter. If you are part of a larger organization, you should have people researching the bleeding edge for possible adoption. That way, you will know when a product can apply to your situation and when it is transitioning from the bleeding edge to the leading edge. We are well past this transition for most GenAI LLM products, such as Copilot 365, ChatGPT, Claude, and Gemini. These products are working today. Your job is to understand how to use them correctly, accurately, and wisely while protecting client confidential information. Beyond my suggestions above, The Fieldguide AI Maturity Framework: The Future of Audit and Advisory is the most thoughtful framework I have seen for thoughtfully transforming your firm from AI novices to leveraging AI strategically. Firms that leverage AI in Assurance will have a clear advantage over those that do not use AI. This framework is highly commendable as it focuses on both technological advancement and human empowerment, ensuring a balanced and sustainable transformation. Your team needs to read, understand, and execute the Fieldguide AI Maturity Framework.
At K2 Enterprises, our commitment lies in providing unwavering support and expert instruction to CPAs. Explore the wealth of resources on our website, where you’ll find valuable insights on selecting the most suitable accounting software, ensuring your firm is equipped with the right tools for the journey ahead. If you work in accounting or finance, K2 Enterprises provides continuing education programs to enhance your skills and credentials. Need help learning how to solve your business’s accounting technology needs and selecting the right software for accounting or CPA Firms? Visit us at k2e.com, where we make sophisticated technology understandable to anyone through our conferences, seminars, or on-demand courses.
