Why AI makes a good companion for tax compliance

Whilst the subject of AI, specifically generative AI, remains one of the hottest topics in the boardroom, it seems the uptake hasn’t quite matched up to the hype. Though tax compliance can be complicated, (so a dose of healthy scepticism is understandable), there are a multitude of compelling cases for generative AI adoption that immediately spring to mind, where this technology would bring immense benefits. Characterised by manually-intensive data entry and processing tasks, and prone to mistakes, tax compliance is an area ripe for the application of AI. It could free staff from much of the daily grind, transforming careers in the tax industry into more interesting and strategic roles. With automation speeding up tedious aspects, tax professionals could apply their intellect to work that relies on human insight and incisiveness, providing much greater value to their organisations and vastly improving job satisfaction.

This is where AI could act as a compliance companion, or co-pilot, not supplanting human intelligence, but carrying out much of the mundane, repetitive work traditionally done by analysts and tax assistants. And there are many great use cases.

Trial balance data mapping 

Trial balance data mapping is still, for most organisations, heavily reliant on the slow manual copying and pasting of data, so mistakes occur frequently. But using AI and automation would radically improve efficiency and accuracy. Automatically mapping raw input from spreadsheets and other documents significantly reduces errors, speeds up the process, and ensures compliance with regulations. In fact, AI tools have been shown to populate the P&L of Corporate Tax returns to an accuracy level of more than 90%, effectively removing around 80% of the time usually spent manually uploading trial balance data into a tax computation. 

Tax professionals could stop performing arduous data entry and instead review ready-prepared data, making any amendments as necessary. Then, generative AI can take the computation another step further by completing around 80% of tax treatment, highlighting the anomalies that need to be addressed. Such a dramatic improvement in the speed and reliability of data extraction, mapping and analysis would be transformational for tax departments, relieving them of the data entry millstone. 

BEPS compliance

Compliance with the OECD’s Base Erosion and Profit Shifting (BEPS) legislation is another serious candidate for AI’s number-crunching capabilities. 

Many organisations have not only had to get to grips with enormous volumes of data to handle BEPS, but also had to retrieve it from multiple, disparate systems located in different countries. Often these systems are running on legacy infrastructures making it complex to assimilate a variety of data formats while at the same time maintaining compliance obligations. Habitually, analysts have turned to Excel to help, but while spreadsheets facilitate consolidation at group level, the regulations stipulate that calculations must also be done at every jurisdictional level, making it far more complicated and time-consuming. 

AI tools easily bridge across siloed systems, identifying and extracting relevant data from various sources, whatever their formats; normalising and processing the data according to BEPS rules to raise any potential red flags. For instance, algorithms can detect unusual transactions between related companies or inconsistencies in transfer pricing that might indicate profit shifting. 

Transfer pricing

Another area in which generative AI can play a key role is within transfer pricing – specifically when it comes to creating benchmarks. To date, this practice has involved junior analysts spending hours and often days, trawling websites to assess whether a company is appropriate for a particular benchmarking exercise. With such an arduous task, it’s no wonder that in practice we see inconsistencies between benchmarks depending on the people and process involved.

Generative AI can remove much of this time and effort; enabling users to review the outputs. So, instead of having to read page after page of website content, taking screenshots and the like, users can instead analyse the results and make their own judgement – being guided by the bot.

Scratching the surface

This is only scratching the surface of how AI could help in the future. It has much more to offer as an intelligent compliance companion, such as monitoring local and global regulations, and alerting firms to relevant changes that need to be incorporated within internal processes. Importantly, AI could analyse existing data, forecast how new regulations would affect tax calculations, identify potential issues in advance and improve forward planning. It would also enable organisations to become audit-ready by organising and classifying data in line with strict requirements, demonstrating transparency throughout the process. This would reduce the risk of non-compliance and potential penalties.

Integrating AI into tax departments will bring many positive benefits for organisations and their employees. Having better tools to carry out mundane work will allow tax professionals to focus on higher value activities and advisory services. For jobs in tax, this offers new opportunities for skill enhancement and faster career progression as professionals develop expertise in managing and interpreting AI systems. In turn, this will create a more dynamic and fulfilling work environment, where tax professionals are eager to take on new technology to drive innovation and work smarter, not harder – as the familiar saying goes.