As organisations continue to explore the opportunities and challenges presented by AI, many tax leaders are facing the same question: where should we focus our efforts to deliver meaningful results?
This was a key theme explored at KPMG's Make the Leap: Tax Technology & AI Summit, which brought together tax, technology and business leaders to discuss the practical realities of AI adoption across the tax function. While the discussions covered a wide range of topics, a consistent message emerged: success depends less on the technology itself and more on identifying the right opportunities, building strong foundations and creating the conditions for sustainable adoption.
Below are six practical steps drawn from the discussions that can help tax leaders turn interest in AI into measurable results.
The fastest teams aren’t chasing novelty, they’re targeting friction. In the “Make AI Happen” breakout, the recommended starting point was to identify the problem and specific challenges, validate them with the team, and only then consider which technology fits, before prioritising with an impact vs effort lens.
A practical way to do this is to build a “use case backlog” across four proven categories discussed in the session:
A recurring Summit “golden thread” was ROI clarity: deep, targeted use cases outperform broad, diffuse efforts, and leaders need a credible business case to scale.
The breakout business case model split value into four components: efficiency gains, cost savings, value generation (better insights/decisions), and governance/control enhancements, supported by a template that forces discipline on current vs future state costs and payback.
One of the most practical sessions addressed a common failure mode: using GenAI everywhere.
The “Beyond AI Hype” content was blunt: use GenAI for messy, unstructured work (extraction, summarisation, drafting narrative, natural language questions) and use rules/ETL/analytics for predictable, high‑accuracy structured processing, especially where compliance requires repeatability and audit trails.
That philosophy was turned into a step‑by‑step decision framework:
Multiple sessions reinforced that AI magnifies what you feed it. good or bad. The Summit whiteboard write‑up called out “data first” as a golden thread: invest in data quality and management before scaling GenAI, and pick the right tool for the job.
Even in the leadership narrative, the point was explicit: start with the data foundation. not the AI, or you will “get wrong answers faster.”
The workforce breakout was equally practical. Its headline takeaway: AI augments, not replaces, and human core skills remain differentiators (decision-making, creativity, accountability). Leaders were urged to set intent, not fear, be explicit that AI is there to support people, and model the behaviour by visibly using AI and agents in their own work.
The disputes/authorities breakout brought urgency with real-world examples of authorities industrialising AI: the Swedish Tax Agency referenced an AI journey that includes policies/guidelines and “around 30 AI‑driven services in production,” plus a compliance case where AI risk selection and automation materially increased audit precision and scale.
For leaders, the implication is practical: prepare for a world where authorities have deeper access to data and more powerful analytics, and ensure your governance and defensibility keep pace.
The Summit’s closing challenge was simple: don’t wait for certainty, build confidence through action. Start with one outcome, choose the right tool, build a defensible business case, and invest in the data and people foundations that make scaling possible.
But ultimately, this is a leadership moment. The organisations making progress are not those with perfect plans, but those with the mindset to act, to test, learn and adapt at pace. Leaders set that tone.
The leap is no longer about ambition. It’s about action.
Discover why organisations across the UK trust KPMG to make the difference and how we can help you to do the same.
