7 Ways AI Can Contribute To The Potential For Error And Overpayment Of Multi-State Sales Taxes

7 Limitations Where Artificial Intelligence Could Affect Multi-State Business Sales Tax Management

PHILADELPHIA, PA, UNITED STATES, November 3, 2025 /EINPresswire.com/ -- An accounting evolution is underway as artificial intelligence is developing for integration into practice, functions and operations. Yet, although artificial intelligence is high on accounting management’s “to-do list,” its actual implementation into accounting practice is still limited as of today. According to surveys, fewer than 20% of accounting firms report actually utilizing artificial intelligence in daily practice. For those firms currently using AI, it is largely confined to functions such as research, data entry, reconciliation and limited forecasting. Why is there such a disparity between interest and implementation?

Multi-state sales tax management is one area of accounting where incorporating artificial intelligence into practice illustrates some of the challenges. William Flick, a thought leader in corporate sales tax policy and process, and a Managing Partner at EisnerAmper, said, “Multi-state sales tax practice is not a tidy endeavor that lends itself to AI enhancement. Because sales tax laws change rapidly, sporadically and often irrationally, coherent correlations and connections can be more difficult to establish. With over 10,000 taxing jurisdictions in the U.S., artificial intelligence in its current form, has the potential to initiate significant errors and overpayments.”

Notes Flick, artificial intelligence is only good as the data behind it. If one’s AI application is based upon a disjointed or disconnected data foundation, it will be challenged in the way it relates to context and the ability to provide meaningful inferences.

Flick points out seven vulnerabilities artificial intelligence can potentially bring to multi-state sales tax management. They include:

1) Limits To AI’s Ability To Understand Sales Tax Nexus.
Artificial intelligence lacks what is called self-awareness and the ability to understand disjointed but contextual meanings. Sales tax nexus is a human construct, made even more convoluted by exceptions and political nuances within each taxing entity. Today’s artificial intelligence iteration will be challenged in making accurate nexus correlations.

2) AI relies on prior data.
Sales tax rates and rules change quickly. In addition, the rapidly changing data, may have little rationale or connection with prior rules and regulations. The body of knowledge available for AI to analyze may be confusing to its making applicable or useful recommendations.

3) AI relies on patterns in data.
Patterns and logic are qualities that the whims of sales tax policy in America often lacks. If sales tax policy has been disjointed in connection to its past, AI will be challenged in drawing logical links to facilitate accurate projections.

4) Product exemptions and taxability from state to state aren’t often consistent or logical.
Product taxability can vary significantly by state (and it often changes frequently, as well). To further complicate matters, sales tax exemptions vary by state on specific products and are often politically motivated and not based on logical projections.

5) AI has been known to make things up if data or connections are missing.
Lacking a proper knowledge base, artificial intelligence has been known to make up its output of recommendations relating to queries. If so, the advice or actions AI recommends can lead to costly sales tax errors and overpayments.

6) Companies using AI can run the risk of compromising corporate privacy.
With AI, data can go in both directions unless protected. Even the best security can leak private data to a public body of knowledge.

7) AI cannot be used in an audit.
State sales tax interest, on the part of companies, is often initiated after receiving a sales tax audit request from the state or municipal government. In a state sales tax audit, artificial intelligence data and its recommendations cannot be presented - one needs original data together with human advocacy and analysis.

Said Flick, “In spite of these limitations, artificial intelligence offers significant potential to enhance efficiency in the practice of sales tax management. Accounting leadership should embrace AI into practice, as the future, where specialists leverage the power of AI as a valuable to support human roles and make practitioners significantly more effective.”
---------

About EisnerAmper Advisory Group
EisnerAmper, one of the largest business consulting groups in the world, is comprised of EisnerAmper LLP, a licensed independent CPA firm that provides client attest services; and EisnerAmper Advisory Group LLC, an alternative practice structure that provides business advisory and non-attest services in accordance with all applicable laws, regulations, standards and codes of conduct. Clients are in all sectors and leverage a complete menu of service offerings. Combined entities include approximately 450 partners and 4,500 employees.

For more information, please contact:
William Flick
EisnerAmper Advisory Group LLC
40 Lloyd Ave.
Suite 308
Malvern, PA 19355
Phone: 484-580-8907
Email: william.flick@eisneramper.com
Website: https://www.eisneramper.com/about-us/professional-directory/bill-flick/
LinkedIn: https://www.linkedin.com/in/williamflick
Twitter: https://www.x.com/BillFlickJr

# # #

Leo Levinson
GroupLevinson Public Relations
+1 215-545-4600
email us here
Visit us on social media:
LinkedIn
X

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.