The regulatory landscape will only grow more complex, forcing organisations to look for smarter, more efficient ways of managing compliance, resulting in the role of AI in tax becoming increasingly more significant.
As part of our Future of Tax series, I sat down with Glenn Price, Global Head of Tax at Vodafone, to gain insight into how this transformation is unfolding in real time for tax professionals. Over our conversation, he shared his perspective on the opportunities, challenges, and future direction of AI in the tax landscape, and how Vodafone has started to implement AI in tax.
Read what Glenn has to say about the rise of AI in tax.
“As AI continues to reshape how organisations function, the tax world is undergoing its own transformation. At Vodafone, I’ve had the privilege of helping lead our adoption of AI across the tax function. While we’re still early in the journey, we’ve already seen significant benefits from boosting team efficiency to improving risk management and enabling more strategic decision-making.
Here’s how we’re currently using AI, where I think we’re heading, and what I believe the next generation of Heads of Tax need to be thinking about.
We’ve made a strong start in using AI within Vodafone’s tax function, although I’d say we’re only scratching the surface of what’s possible. There are five key areas where we’re already seeing tangible impact.
To begin with, everyone in the tax team now has access to tools like Microsoft Copilot, and we actively encourage exploration. Giving the whole team access has democratised innovation, allowing people at all levels to experiment and bring new ideas to life.
We’re also using AI to streamline tax data extraction and classification. For example, we use Copilot in Excel to analyse data and improve our search terms. If we’re looking for disallowable legal or professional expenses, we don’t just search manually – we ask Copilot to expand and refine our search parameters, which leads to better, more accurate results. It’s helping us search smarter and faster.
One of the most exciting developments has been in predictive analytics. We’ve applied machine learning to forecast tax liabilities, using confidence scoring to assess data quality within tax returns. When something appears unusually high or low, we can drill down into the data quickly and get to the root of the low confidence level much faster. What started as a tool to support our compliance work is now being applied in new areas, which could traditionally be hampered by unfamiliar datasets.
Another area I’m particularly enthusiastic about is the Copilot Researcher agent. With the right prompt, it can produce a first-cut technical tax analysis that can be impressively close to what a senior manager might deliver. Whilst it’s not a substitute for human expertise, it’s a powerful tool that accelerates our initial thinking and supports more informed decision-making. The ability to generate such a comprehensive analysis in 15 minutes is remarkable – not perfect, of course, but hugely helpful as a starting point. We have one project where I estimate that Copilot Researcher accelerated our work by about one week.
Finally, we’ve been running training sessions across the team, not just to teach the tools but to share and listen to early success stories. Seeing how peers have used AI in practical scenarios is motivating others to try it for themselves. That peer-to-peer learning has been just as valuable as any formal training.
Although we’ve made solid progress, it’s clear we’re still at the beginning of our AI journey. What excites me most is the potential for even deeper integration between tax and the wider organisation.
We’ve already started building those bridges. Our tax technology specialist has spent time embedded with the data and analytics team, learning their tools and methodologies so we can speak the same language. This alignment has been crucial, especially as Vodafone’s technology team is driving consistent use and governance of AI across finance. In fact, I’m running a session with our finance leadership team later this year to share progress across our different teams.
Looking further ahead, machine-to-machine communication is both here and on the horizon – it’s both fascinating and a little daunting. As more tax authorities adopt digital tools and real-time reporting, we’ll start to see tax authorities increasingly adopting digital tools and real-time reporting with increased automation, particularly in the early stages of audits. This will require robust governance, clean data, and transparent outputs – areas where Vodafone is investing to ensure responsible and compliant use of AI.
That, in turn, changes the role of the tax professional. Tools like Copilot Researcher are already capable of reading legislation, case law, and tax authority guidance – and can provide high-quality summaries and even confidence scoring. A year ago, I wouldn’t have imagined that. Now, I find myself reviewing AI-generated analyses that might be between 90–98% accurate.
The role we play is shifting towards human judgement – interpreting the data, challenging prompts and assumptions, and validating whether we’re comfortable with the conclusions AI has drawn. We’ll still need people, but the work that some of us do has – and will continue to – change.
And that also changes how we work with our external advisors. I can foresee a shift in how in-house tax teams and external advisors collaborate – with in-house tax teams increasingly equipped to conduct detailed tax technical analysis supported by AI, and advisors providing strategic validation and specialised insight. That could be a fundamental shift in the client–advisor relationship, and I feel it’s coming faster than many expect. Advisors could struggle in such a world if they can’t provide specialised or strategic insight.
The most important message I’d share with future tax leaders is this: embrace the technology and do it now. This is not about your age or how technical you are – it’s about mindset. I’ve seen colleagues in their 50s (like me, just!) experiment with AI far more than junior team members. It’s the curious, growth-mindset individuals who are thriving. But we have to start experimenting with our own prompting and learning.
At Vodafone, we’ve supported learning through our internal training platform and by encouraging informal exploration. I’ve personally spent time on YouTube to see how others prompt, experimented with prompt coaching, and used Copilot agents to build my understanding. You don’t need to know how the model works; you just need to know how to ask it the right questions.
A colleague of mine has been brilliant at this. She’s constantly asking, “Can we use AI to do this?”, and together as a team, we figure out that we can. You don’t need to be a computing expert to benefit from AI – what matters most is knowing how to ask/prompt the right questions and being open to experimentation. A foundational understanding of the tools is helpful, but curiosity and problem-solving are the real drivers of success.
One tip I’ve picked up: after using several prompts to reach the outcome you want, go back and ask Copilot to write a single prompt that would have got you to the final output directly. I’ve done this for a reporting task with 15 prompts, and now I have a reusable shortcut that saves huge amounts of time.
Looking ahead, I think the skills that will matter most for in-house tax teams are not an encyclopaedic knowledge of technical tax rules but the ability to spot efficiencies, ask good questions, and use AI as a partner to ensure simplicity and efficiency. Everyone – whether they’re tax professionals, data experts, or junior analysts – needs to lean into this. Those who embrace AI now will be well-positioned for the future. The pace of change is rapid, and engaging with these tools early offers a valuable opportunity to stay ahead and shape how the profession evolves.
AI is already reshaping the tax function, and it’s only accelerating. At Vodafone, we’re seeing real improvements in efficiency and accuracy in tax compliance, reporting, and risk management. But the most exciting part is what lies ahead.
As tax becomes more interconnected with technology and data, we need to evolve. That means being curious, embracing change, and building new skills. The traditional training path – learning through compliance and spreadsheets – will be complemented (not replaced) by new ways of working, where technology does the heavy lifting and humans focus on insight and judgement.
I believe those of us who lean in now will be at a real advantage in the years to come. And for those entering the profession, there has never been a more exciting time to be in tax. Anyone asked to conduct some fairly laborious tax work can probably now use AI to get it done smarter, faster, and better.”
If you would like to take part in the Future of Tax discussion series, get in touch and let’s explore a conversation from your perspective.
