Organisations are investing in AI at a remarkable rate. KPMG's new Global AI Pulse report finds companies planning to invest $186 million on average in AI over the next year.
But investment is one thing; return is another. Only 8% of 2,000-plus business leaders say they’ve achieved an ROI on their AI solutions.
The report is clear about why: what’s missing is not access to the technology, but the ability to deploy it at scale. And the key to that ability is governance.
Our survey found that 81% of AI leaders – those firms delivering business value from their AI investments – have the governance in place to manage AI risk at scale. That compares with 63% of non-leaders.
Why is governance so crucial to AI adoption? How do you embed it into your tax operations, and what challenges will you face along the way?
AI governance: From oversight to enabler
Robust governance is what allows organisations to scale AI deployment.
It’s easy to perceive governance as a handbrake on technology adoption: a layer of oversight and control that slows things down, and limits what you can do. In fact, the opposite is the case. Without governance, AI doesn’t scale. It stalls, breaks or creates risks that force organisations to pull back.
In tax environments, the cost of failure isn’t just operational; it’s also regulatory. Errors in AI-generated filings can trigger audit scrutiny, penalties and reputational risk. Good governance helps you prevent mistakes, and catch them early when they do. It releases the handbrake.
AI governance builds trust with stakeholders
It also builds trust in the solutions you're using among internal and external stakeholders.
Internally, your team needs to feel comfortable operating and overseeing AI tax tools, and confident that they’re not creating risks for themselves or the company.
Externally, meanwhile, tax authorities are paying close attention to how organisations use AI. They want to know whether they can rely on AI-generated outputs from taxpayers, or whether they will need to verify them. Being able to demonstrate effective governance will give them confidence in the numbers you report to them.
What’s more, strong AI governance helps you identify the right use cases.
There’s a temptation to employ AI tools simply because they're available. By following a defined governance process, tax teams can evaluate the benefits, scalability, risks and costs of potential use cases, so that you implement the ones that will genuinely add value.
Embedding AI governance in tax operations
As our report underlines, embedding AI governance into your tax operations means aligning it with your systems and workflows.
That’s crucial as you scale AI, because it reduces duplication, inconsistencies and cost, while accelerating adoption.
The trick is not to treat governance as a separate layer, sitting on top of existing processes. Build your controls directly into the processes they’re meant to support, so that they’re business as usual for your team members.
When governance happens in the flow of work, people are far more likely to follow it. Parallel processes, disconnected from the tasks at hand, risk not being carried out.
Ensuring consistency is also vital – especially in complex, decentralised businesses. Your AI governance policies should be defined centrally, then clearly articulated throughout the organisation. And they should include measures to escalate any issues that crop up.
Integrating AI governance into the tax function isn’t a once-and-done exercise. You’ll need to make sure it keeps pace with the rate of technological change.
AI is evolving faster than regulatory frameworks can account for. And governments are taking different approaches – some prescriptive, some principles-based. Waiting for the legislative landscape to become clear isn’t an option.
Tax leaders must instead look to internal governance. Make sure you – or somebody in your team – can monitor how AI is being adopted within the wider organisation, and update your guardrails accordingly. That may be a dedicated AI governance role, a tax technology specialist, or process owners across VAT, transfer pricing and corporation tax.
Start with a clear view of your AI landscape
An essential first step, before you begin putting your AI governance in place, is to know what solutions the tax and finance functions are using. If you haven’t already, build an AI inventory that sets out:
As AI systems proliferate – rapidly – they’ll become ever harder to take stock of. Doing so now will avoid trying to play catch-up further down the line.
Our experts can help you to create this. KPMG's Trusted AI Framework provides a foundation for safe AI adoption, based on core principles such as explainability, transparency, accountability and data quality.
We’ve applied these to the specifics of the tax function, giving you a systematic approach that turns governance from an oversight mechanism to a driver of AI adoption.
Tax functions that treat governance as a constraint will struggle to scale AI. Those that use it as an enabler will define the future of the function. Get in touch to see how we can support your approach to AI governance in tax.
Our tax insights
Something went wrong
Oops!! Something went wrong, please try again
Get in touch
Discover why organisations across the UK trust KPMG to make the difference and how we can help you to do the same.