Why AI in tax is no longer optional

So as the work itself becomes more complex and more interconnected across jurisdictions, there aren’t enough people to handle all this change. And there’s less and less room for error.

For Ray Grove, Global Head of Product for Tax and Trade at Thomson Reuters, this isn't a future scenario. It's the reality tax teams are navigating right now. "Getting it good isn't good enough," Grove says. "You have to get it right."

That tension between precision and scale is exactly why AI in tax has moved from optional to essential. Automation alone can't substitute judgement. But without it, and without AI to make sense of the data behind every decision, tax functions simply can't keep up.

Tax operates in a high-stakes environment in which errors are not just inconvenient; they carry regulatory, financial, and reputational consequences. At the same time, tax professionals know meeting regulations is far from simple.

“You might look at tax and think it’s black and white,” Grove explains. “But there’s a whole lot of grey.”

The combination of set rules and real-world nuance makes tax one of the most demanding use cases for AI. Systems must process large volumes of structured data, interpret complex regulations, and support decisions that require professional judgement.

All this in addition to the fact that output needs to be accurate, auditable and defensible.

It’s common to frame tax as a proving ground for emerging technology. But Grove sees it differently.

“I don’t know if it’s a proving ground as much as a necessity for how the future is going to work,” Grove says.

Three main forces are driving that necessity:

Businesses operate across multiple jurisdictions, each with its own requirements, timelines, and interpretations. As Grove puts it, compliance is “one of the most fascinating and interconnected problems that exists.” It determines everything from who companies can do business with to where they can operate.

Despite the automation potential, Grove is clear that AI is not about removing the role of the tax professional. Instead, it changes the nature of their work.

The opportunity, he suggests, is to move away from manual, data-heavy tasks and toward higher-value decision making.

AI can “help the subject matter expert go away from being a subject matter expert and really focus on the judgement decision that I’m making,” he explains.

But Grove is not talking about just any AI. In high-stakes professional work, the source of the answer matters. “When you look at general-purpose AI, like just basic Claude or basic ChatGPT, it’s grounded on the internet,” he says. “I don’t know about you, but I’ve seen a few things on the Internet that strike me as suspicious.”

That is the distinction behind Thomson Reuters’ Fiduciary-Grade AI™: AI built for professional environments, grounded in trusted, curated legal and tax content, and shaped by people who understand the standards tax professionals are expected to meet. 

In practice, that means using AI to analyse data, identify patterns and surface options, while keeping humans firmly in control of final decisions.

It is a shift from execution to oversight, from processing to interpretation.

It’s not just businesses that are evolving. Tax authorities are also becoming more sophisticated in how they use data and technology. Meaning, the regimes have access to advance AI tools, too.

Authorities already have “an incredible amount of data with a very clear taxonomy,” Grove points out. And they are using it to improve how they identify risks and enforce compliance.

That creates a new expectation: companies must operate with the same level of intelligence and visibility as governments do.

“I think they kind of expect that you’re going to do the same,” he adds.

In other words, staying manual is not a neutral choice. It’s one that introduces internal and external risks, potentially putting businesses at a disadvantage.

Tax isn't slowing down. Transaction volumes will continue to grow. Real-time reporting requirements will expand. And the talent gap isn't closing on its own.

The organizations that move now to integrate AI and automation into their tax operations won't just be more efficient. They'll be more accurate and better positioned to stay compliant and competitive.

As Grove puts it, "I think the biggest risk is not moving."

For tax leaders, that insight is hard to ignore.