Overcoming generative AI risks within tax
With its power to transform the way professionals work across all industries, generative AI has exploded onto the scene with astonishing capabilities that mimic human-like intelligence.
Jump to:
With its power to transform the way professionals work across all industries, generative AI has exploded onto the scene with astonishing capabilities that mimic human-like intelligence. However, as the saying goes, with great power comes great responsibility.
From an accountant’s perspective, there is no question that the benefits of AI are significant, it’s important to proceed with caution. As open-source generative AI platforms (like ChatGPT) gain momentum, there are concerns around accuracy, security, and privacy — especially when it comes to handling clients’ sensitive tax information.
To mitigate these concerns, forward-looking accountants are relying on AI-powered tax research solutions that are grounded in data from trusted experts and reliable sources, as opposed to public options that pull information from across the internet.
But before we jump into how accountants can overcome the risks of generative AI, let’s take a look at the basics of this revolutionary technology.
The basics of generative AI in tax
So, what exactly is generative AI and how does it work within the tax and accounting profession? In short, generative AI refers to a class of artificial intelligence systems designed to generate new and original content, often mimicking human-like capabilities. Unlike traditional AI models that rely on pre-programmed responses or patterns learned from data, generative AI has the ability to create novel outputs independently.
An open-source and publicly available generative AI platform, like ChatGPT, is trained on vast datasets from across the internet. Once trained, these generative AI models can produce content, such as text, images, or even music, which wasn’t explicitly present in the training data, but was learned from it.
From an accountant’s perspective, generative AI opens the door to much more streamlined and highly efficient processes, particularly when it comes to tax research as it can provide instant access to insight on tax laws and analysis.
The risks of generative AI in tax
While the benefits are significant, there are also concerns regarding the security and privacy of sensitive information, prompting the need for careful consideration and communication with clients on the use of generative AI.
But what exactly are the risks? To start, the Large Language Models (LLMs) inherent in generative AI are trained on large amounts of data, however, we don’t know if that data is confidential, proprietary, or quite frankly, accurate. Because tax professionals rely on accuracy and preciseness, there is a concern when using publicly available generative AI models for professional purposes as there is no way of knowing what pieces of data the model based its responses on – and no way to validate them.
In addition, a phenomenon known as “hallucinations” can occur when AI models fabricate a confident but inaccurate response. This issue can be caused by a number of factors, including divergences in the source content when the data set is incredibly vast, or flaws with how the model is trained. The latter can even cause a model to reinforce an inaccurate conclusion with its own previous responses. Clearly, this is an unacceptable risk for tax and accounting professionals.
Understanding client concerns about generative AI
From a client standpoint, concerns about identity theft and unauthorized access to sensitive information are valid when it comes to open-source generative AI platforms. Clients may worry about the possibility of their financial information being uploaded into generative AI systems.
Data breaches and unauthorized access are real risks, and acknowledging these concerns is the first step toward choosing professional AI-powered tax research solutions built specifically for tax professionals and backed by the highest standards in security.
Building trust in generative AI
Transparency and accountability play pivotal roles in alleviating client concerns. Accounting firms must communicate the steps taken to ensure the security and privacy of client data.
Emphasizing encryption, secure data storage, and access controls, while adopting industry best practices and certifications, can significantly mitigate risks and instill trust in the use of generative AI as part of your firm’s daily workflow.
Remind your clients that your firm is available to answer any tax-related questions and that the generative AI solutions available to tax and accounting providers (i.e., Checkpoint Edge) are much more secure as they pull answers only from the most trusted experts and vetted sources.
Educating clients on generative AI
Clients’ discomfort with AI, regardless of its security, can be addressed through education.
Explaining the benefits and limitations of generative AI in tax procedures, along with detailing the safeguards your firm has in place, is crucial. It’s also extremely important to advise your clients not to copy and paste or upload any of their personal tax or financial data to open-source generative AI platforms in search of answers. Make sure they come to you as a first point of contact.
Demonstrating the value of generative AI
The potential benefits of generative AI within accounting firms can be transformative — if a professional platform built specifically for tax and accounting professionals is used. Increased accuracy, efficiency, and compliance are just some of the tangible advantages.
Here are several ways generative AI can transform accountants’ daily workflow:
- Tax research and compliance: Generative AI can assist in keeping track of evolving regulatory requirements and ensuring that financial practices comply with the latest standards. This helps in reducing the risk of non-compliance.
- Data entry automation: Generative AI systems can automate data entry tasks by extracting information from various documents, such as invoices and receipts. This can reduce the manual effort required for data input while minimizing errors.
- Document generation and summarization: Generative AI models have the potential to automatically generate financial reports, invoices, or other accounting documents. They could also assist in summarizing lengthy financial statements or reports for quick insights.
- Fraud detection: AI, including generative models, can be used to identify patterns and anomalies in financial data that may indicate fraudulent activities. By analyzing large datasets, AI systems can help auditors and accountants detect irregularities.
- Forecasting and predictive analytics: Generative AI models can be applied for predictive analysis by analyzing historical financial data. This can help accountants forecast future trends and make more informed decisions.
- Natural Language Processing (NLP) for client communication: NLP, a subset of AI, can be integrated into communication tools to enhance interactions with clients, create automated responses to common queries, or even assist in drafting emails and reports.
Making generative AI work for you
At the end of the day, overcoming client concerns about generative AI in the tax industry requires a multi-faceted approach. By understanding, addressing, and actively mitigating the risks of generative AI, your firm can build trust and educate clients on the benefits of this transformative technology—while also using it to your competitive advantage.
By adopting an AI platform that is grounded in expert data from reliable sources, your firm can ensure precision, accuracy, and efficiency in your tax research and beyond, while better serving your clients.
For more insight on generative AI, we invite readers to subscribe to our Tax Professionals Newsletter and check out our recent webinar below:
NewsletterTop Reads – Tax Professionals Join us each month as we explore the role and impact of AI on tax, accounting, and audit professionals. Learn how to apply AI ethically and responsibly in your work, and what skills you need to thrive in this AI era. |
