The international tax landscape changed forever in March 2023. Co-founder & President of OpenAI, Greg Brockman, input the United States Tax Code into ChatGPT, which then calculated a hypothetical couple’s tax liability for the year. Even Brockman himself seemed surprised by the technology’s prowess.
“It’s an amplifying tool,” he said, summarising ChatGPT, “that will let you reach new heights.”
The word amplifying may have seemed contradictory to those who believed that Brockman’s demonstration signalled the end of the traditional tax filing process.
Currently, there is a glut of online content exploring tentative use cases for AI in tax and audit contexts. The reality is that, for many tax & audit firms, AI is on the radar, but hasn’t been fully embraced. Indeed, it may be challenging to imagine how AI’s polymathic abilities might ‘fit in’ to the complex, cross-silo interactions within large tax and audit firms.
In this article, we’ll discuss a proven use case for AI for tax and audit processes – data extraction.
AI data extraction describes capturing relevant data points across various documents. Over the last few years, AI data extraction has evolved from a frail, inconsistent technological framework into a powerful tool that many companies rely on to capture data:
Tax preparation and audit processes rely on carefully capturing and organising information. A study by PwC shows that tax teams spend 40-60% of their time on data management tasks and only 30% on strategy-based initiatives. The operational costs of completing a routine task, like data extraction, inhibit long-term growth. Or, to put it simply – using employees to extract data isn’t scalable.
AI-automated data extraction offers more benefits than simply scraping data from a few financial statements. As KPMG notes, AI can assist auditors in monitoring risks in larger and more complex datasets than is realistic using manual analysis.
One valuable ability of new AI data capture technologies is capturing and structuring the relevant data as soon as the user inputs the document i.e., real-time extraction.
Access to cleansed data significantly enables tax and audit organisations to quickly identify any inconsistencies and errors at an early stage. A proactive approach to anomaly detection is crucial in the highly regulated and detail-oriented fields of tax and auditing.
Moreover, the capability to extract data in real-time also significantly enhances the speed of communication between organisations and their clients. Such improvements in efficiency and responsiveness foster better working relationships between clients and tax and audit organisations.
Thoroughly reviewing past records can reveal specific overlooked deductions and strategies for legally reducing taxable income. For example, analysing a client's historical tax returns can pinpoint precise areas for cost savings, and mechanisms for income exemption.
If your organisation is not seeking a permanent AI data extraction solution, you might consider using a vendor to extract the data from a large batch of files (e.g., a backlog of tax returns, financial statements, bank statements, contracts, etc.) as a one-time solution.
The AI data extraction market is full of vendors from different countries with diverse specialities. Finding a provider with scalable, robust data extraction technology is a cumbersome task. But it doesn’t have to be if you know what to look for in an IDP tool.
Introducing AI into a tax and audit context may not be met with open arms. Despite the repeated assurances from major tax & audit firms about their commitment to developing AI frameworks, the reality is that employees can be antagonistic towards implementing AI. Fear Of Becoming Obsolete (FOBO) escalated amongst college-educated workers in 2023. As a result, employers may have to adopt new strategies to combat stress, paranoia, fear and resentment among their employees.
FOBO isn’t typically an issue when it comes to data extraction, though. As our clients have noted, no one seems to like manual data entry, anyway.
However, in 2024, the word ‘AI’ comes with connotations, making it vital to outline the aims of any AI initiative. Implementing AI to extract data from invoices and receipts sets clear expectations for employees. On the other hand, simply announcing that AI will be introduced into the firm’s tax & audit processes doesn’t.
Relinquishing customer data to a third-party vendor may also be a source of anxiety for companies. Corporate tax is a tightly regulated industry, and handling complex client data requires rigorous security measures. However, embedding technology in your company’s IT infrastructure doesn’t constitute a security risk if the technology deploys specific security measures (such as having an ISO 27001 certificate, encrypting data in transit and conducting regular penetration tests).
Find out more here.
In 1970, American computer scientist, Marvin Minsky, told Life Magazine, “From three to eight years we will have a machine with the general intelligence of an average human being.” Though his prediction was overambitious, he correctly anticipated a technology with advanced natural language processing (NLP) capabilities and the ability to learn from its mistakes.
Going forward, AI will become even faster, more skilled and cost-friendly. Not to mention, more widely adopted by tax & audit firms around the world.
AI-powered automation offers a variety of benefits in tax and audit contexts. Of course, AI can’t replicate the insights of an accountant with years of experience. But what it can do is extract pertinent information from a variety of financial documents, easing your team’s workload and supporting better decision-making.
Interested in learning more? Book a demo or email us at hello@evolution.ai.
