The tipping point: Measuring the success of AI in tax

Since our last survey in May 2024, the percentage of tax professionals using generative AI for work purposes has jumped from 35% to 40%. Intent to use AI has risen significantly over the same time period, from 32% to 49%. The percentage of tax practitioners with no plans to adopt AI shrank dramatically, falling from 33% to only 8%.

89% of tax practitioners are either using or planning to use generative AI

Tax and technology speaker and Chartered Institute of Taxation (CIOT) Vice President, Paul Aplin, says: "Generative AI hit the headlines with a bang in late 2022. Some thought it would revolutionise the profession while others thought it would never deliver on its promise in a professional context. The evidence in this survey is clear: AI is now a standard tool in the tax practitioner's toolbox."

For tax practitioners, AI means staying ahead, being more efficient, and providing a better service to clients, says Ian Bowden, Tax Partner at BDO.

The key drivers for using generative AI, according to the survey, are delivering work faster (74%), a better service to clients and stakeholders (53%), and having a competitive advantage (52%).

Three-quarters of tax professionals say the key benefit of generative AI is delivering work faster

Despite this huge majority, nearly half of tax professionals feel their firms are slow to act.

Sign up to the  Tolley+ AI™ Insider Programme  to get updates and breaking news on the latest AI developments in tax. 

Private practices and in-house teams have been relatively quick to evaluate, implement and integrate AI into their tax services, but practitioners are eager for their organisations to continue pushing for greater AI-powered innovation, and faster.

Our survey found 48% of tax practitioners believe their organisation is slow or very slow at implementing new technology, especially AI.

Half of tax practitioners said their organisation is slow or very slow at implementing new technology

This is the opposite of what you would have expected in the past, says Georgiana Head, the Director of Georgiana Head Tax Recruitment, when practitioners feared the loss of jobs because of automation of compliance by AI.

Jane MacKay, Tax Partner at Crowe, says her team has a group of very enthusiastic tax professionals – across all grades and career stages – who are embracing the changes brought on by AI and taking the time to upskill.

However, this is not the case for all practitioners. “There are also a lot of busy tax professionals who struggle to make the time to experiment with AI and other tools,” she notes.

Find out more about Tolley+ AI

MacKay hopes the tax profession's active engagement in AI has reached a “tipping point” – a reference to Malcolm Gladwell’s book The Tipping Point – which should give momentum to wider implementation.

A common barrier to change or innovation is comfort, says Hayley McKelvey, the Digital Innovation Leader for Tax & Legal and a Partner at Deloitte.

When asked what would encourage them to use AI more often, the responses show a clear appetite for AI advancement. Almost two-thirds (60%) said they want more training, while 47% said better communication around use guidelines. Other commonly expressed answers were a desire for further communication around how AI works (40%) and easier integration with existing workflows (40%).

60% of tax practitioners said they'd use AI if they had more training

AI fluency programmes are key to increasing tax practitioners' core understanding of AI and will boost adoption rates, says McKelvey.

"This isn’t about turning our tax practitioners into data scientists, but it is about building a high level of understanding that promotes confidence and trust.”

How firms are implementing AI

Ensuring AI adoption isn’t just about initial enthusiasm; it requires long-term leadership and reinforcement. Leaders must help their teams navigate AI adoption in a way that is sustainable and supported, says McKelvey.

Although there is a clear desire within the tax profession to expedite the integration of AI technologies into tax workflows. Jonathan Scriven, Director of Tax Markets at LexisNexis, says it is evident that many businesses are encountering challenges with implementing technology at speed.

While most tax professionals are using or planning to use AI, the demand for more is only continuing to accelerate. But adoption alone is not enough. Firms must integrate AI into their workflows in a way that delivers tangible results.

The top steps organisation have taken to implement AI include offering an AI powered product to staff (36%), training staff to use AI (18%) and developing generative AI policies (17%).

A third of respondents said their organisation had launched an AI powered product to staff

"I suspect closer to 100% of tax practices actually use some form of AI product," says Head from Georgiana Head Tax Recruitment.

"People just don’t realise they're using it. Anyone who uses Teams for their meetings could have access to co-pilot and any practice which uses Xero or similar accounting software will be using some form of AI."

While organisations are experimenting with AI, structured implementation remains slow and many lack clear strategies for how to embed AI into their workflows effectively.

Demand for faster implementation is also high amongst in-house tax teams. A financial controller at a furniture manufacturer says: "Practices that do not embrace AI will soon find themselves behind the curve and at a competitive disadvantage as this type of technology becomes more normal."

The first tax practices to take AI's potential, reduce time spent researching or drafting, then lower fees to clients will undoubtedly reap the rewards, says a Director of Tax at a major leisure and hospitality company.

With nearly 90% of tax practitioners either using or planning to use AI, firms must move quickly beyond basic experimentation. The real differentiator will not be whether firms adopt AI, but how effectively they embed and measure it.

Failing to act on AI isn’t just a missed opportunity, it’s a serious business risk. Firms that lag behind won’t just lose efficiency. They’ll lose their workforce.

The data suggests we are at a tipping point. Tax professionals recognise the value of AI, but if they feel their firms are holding them back from working more efficiently, they may look elsewhere for opportunities. : Firms that fail to invest in AI not only risk inefficiency but also losing their workforce.

1 in 5 tax practitioners will consider leaving their firm if it doesn’t invest in AI

The survey revealed the impact that slow or no AI investment would have on the careers of practitioners. More than a third (35%) said their career would be negatively impacted if their organisation failed to embrace AI, while one in five (20%) said they would consider leaving if their organisation failed to embrace AI.

This means AI investment is not just about improving efficiency. It’s about retaining talent in an increasingly tech-driven profession.

"Recruiting high-calibre people has always been challenging and it is crystal clear that the best candidates will gravitate to firms that embrace AI," says Aplin from the CIOT. 

Bowden from BDO argues that firms must overcome their fears by considering how AI enhances, rather than replaces human skills. 

Firms that act now will build a strong AI-powered workforce. Those that hesitate risk talent drain and competitive decline.

Transform your tax workflow today with Tolley+ AI

Despite the clear benefits of AI, firms are failing to prove its success. If they can’t demonstrate ROI, investment will stall and adoption will slow.

Tax practitioners view the top benefit of generative AI as increased efficiency. This has a far-reaching knock-on effect on most aspects of a tax practice or finance department.

However, attributing a precise metric to time saved, risk reduced, or the impact on employees is not simple.

The most commonly used metrics to measure the impact of AI are time-savings (41%), accuracy of outputs (26%), cost savings (21%) and employee feedback (13%). Almost half (42%) of tax practitioners said their organisation isn’t looking at any of the obvious metrics.

Almost half of tax experts say their organisations aren't using any of the obvious success metrics

For us, success isn’t just about time or cost savings, says Bowden at BDO.

Keeping an eye on traditional metrics, such as return on investment and efficiencies, is essential, says McKelvey from Deloitte, but organisations also need to take an holistic approach when assessing performance.

Whilst we might default to the traditional metrics, she says, there is also a need to take a longer-term view given the profound impact that AI will have on our business.

"Some tools and techniques deployed in the business will not deliver traditional business value, certainly not in the short term,” continues McKelvey.

“Instead, they might represent a strategic deployment to help acclimatise our people to working more effectively with AI and in a way that creates trust.”

In the longer term, this helps ensure the successful adoption of more sophisticated AI tools, she says.

One in-house tax leader said their department measures AI's effectiveness by evaluating through accuracy, efficiency, and user adoption rates.

"For tax professionals, key metrics include error reduction, time savings, and compliance accuracy."

Another in-house tax leader said their organisation is currently in the early stages of identifying best practices, devising organisational policies, and determining best providers.

"The most important metrics are time saved in compliance work and in research. A good example being AI tools that browse libraries of verified tax legislation and guidance to compile advice."

While measuring AI’s impact is essential, firms must also address the risks that come with adoption. Without clear safeguards, AI could create new challenges around data security, transparency and over-reliance.

While AI presents significant opportunities for tax professionals, it also brings inherent risks. Firms that fail to address these risks may struggle to build trust, both internally among employees and externally with clients.

Hallucinations were the top concern (60%) among tax experts, followed by over-reliance on AI (59%) and leaking confidential data (43%).

Hallucinations are the top concern for tax professionals when using AI

At the same time, tax professionals are open to using AI-powered tools if they are grounded in trusted content.

Almost three-quarters (72%) of respondents said they would be somewhat or completely confident using AI-powered tools grounded in tax content sources.

There is, unsurprisingly, still concern over accuracy of output, says Aplin from the CIOT.

"Quality and reliability will improve as firms hone their prompt engineering skills. Investment in training, the establishment of clear guidance and strong governance will all help to reduce the risks of using generative AI. If people know where the guardrails are they will be more inclined to experiment and push the boundaries of the technology."

A quarter century ago people said that search engines were not appropriate tools for use by professionals researching tax problems, says Aplin.

"I never bought that idea: knowing which sources you can and cannot rely on is a basic professional skill and it is as applicable to the output from AI Large language Models trained on tax material."

Transparency with clients is also an important way to reduce risk. In-house tax practitioners expressed a strong demand to be informed if the external advisers who they are working with had used AI. More than two-thirds (68%) of in-house experts expect to be informed, compared to just half (51%) of private practice professionals who think their clients want to be informed.

Most in-house tax leaders want to know when their external firms are using AI

Responsible AI adoption will be a key differentiator for firms. Tolley is part of RELX, a global provider of information-based analytics and decision tools for professional and business customers, which has clearly established responsible AI principles in place.

"As part of our responsible approach to developing AI solutions, we always consider the real-world impact of the solution," says Scriven.

Sign up to the Tolley+ AI™ Insider Programme to get updates, early feature access and breaking news on the latest AI developments in tax.