2023 and 2024 were turbulent years for a research and development (R&D) claimant. HMRC’s compliance activity was at its peak, with the task force intervention resulting in a seismic increase in enquiries into R&D claims. Perceived as the highest risk group for fraud or system abuse, small and medium sized enterprises (SMEs) bore the brunt.

By 2025, the pressure had eased. Enquiries became better targeted. Dialogue with HMRC grew more constructive and cases began reaching sensible conclusions. The aftermath of the quake was a displacement of the profile of claims. SME claims fell by 29%, countered by a 36% increase in the amount claimed by large businesses.

It would be easy, by comparison, to see the current status quo of the R&D regimes as the calm that follows the storm – but beneath still waters, strong currents of complexity, and subjectivity remain, which can have a material impact on a company’s R&D claim.

The territorial shift: where your R&D happens now matters

For accounting periods beginning after 1 April 2024, the UK R&D regime moved to a largely territorial basis of claim, mirroring many other comparable regimes globally. The intent was clear: incentivise innovation that stays on UK soil.

Businesses using overseas workers (not on a UK payroll), or contracting R&D activities to overseas companies generally couldn’t include these costs in a UK R&D claim. Broadly speaking, overseas costs can only be claimed, where:

• The activity was necessary for the purposes of R&D;
• The capability didn’t exist in the UK;
• The capability was present in the overseas location; and
• It would be wholly unreasonable for the company to replicate these conditions in the UK.

Armed with this guidance, and extensive examples in HMRC’s own manuals, businesses must apply these conditions to their existing commercial arrangements, and for each, assess the reasons behind their use of a particular overseas resource (be it labour, or contracted out activity).

That fourth condition is where it gets difficult. “Wholly unreasonable” is not the same as “quite difficult” or “commercially inconvenient”. The bar is high and the language is inherently subjective.

Businesses must apply these tests to their existing commercial arrangements, case by case. And critically, they must be able to evidence not just what they did, but why they didn’t use a UK alternative.

Take a business that contracts with a European specialist test house or laboratory to run tests on its behalf. Before including these costs in a UK R&D claim, they will need to evidence that an equivalent UK capability either doesn’t exist or couldn’t reasonably be used.

These costs can often be significant, and span industries. In the technology and media sector, for example, offshoring software engineering is commonplace. In life sciences and healthcare, businesses will find the right skillset, wherever it may be. If a world-renowned expert in a particular condition is based at Harvard University, then that is who the company will use.

Whether those costs qualify under the new rules is a question that demands careful, expert analysis.

R&D tax claims in complex supply chains: who is entitled to claim?

Also effective for accounting periods beginning on or after 1 April 2024, were new rules designed to prevent multiple parties claiming R&D relief on the same expenditure in complex supply chains.

This was intended to be achieved by introducing the concept of “contracted out” R&D.

The principle is straightforward enough. Where Company A engages with company B, R&D is contracted out where:

• Company A enters into a contract;
• The activities performed to meet the contractual obligations include R&D; and
• It is reasonable to assume that Company A intended or contemplated that R&D activity would be required in performance of the contract.

Where these conditions are met, Company A may include the relevant costs of this contracted out R&D in its own R&D claim. Company B cannot.

In contrast, if these conditions aren’t met, then it is possible that the R&D claim may be made by Company B (where it can demonstrate that R&D has not, in turn, been contracted out to it).

Again, we are faced with subjective language – “intended or contemplated, having regard to the terms of the contract and any surrounding circumstances”.

Businesses and their advisers must carefully consider the evidence, the intentions and any other records kept when assessing where the appropriate R&D claim resides. In some cases, it will be clear cut; but for many complex supply chains, we are faced with another complex, judgemental area.

Using AI in R&D tax claims: powerful tool, real risk

No sector is untouched by the rapid rise of AI and R&D advisory is no exception.

The potential is real. AI tools can improve the quality, consistency and presentation of R&D claims significantly. Used well, they make the process more rigorous and the output more compelling.

But used poorly, the risks are equally significant. An over-reliance on automation, by advisers, by claimants and potentially by HMRC in its assessment of submissions, could erode the quality of judgement that underpins a well-constructed claim. R&D tax relief is not a form-filling exercise. It is a technical, evidence-based assessment that demands qualified, experienced human oversight at every stage.

Navigating the road ahead

The calmer R&D waters of 2026 are welcome. Companies, advisers and HMRC are now returning to a more collaborative, engaged approach to making, reviewing and agreeing R&D claims. But calmness shouldn’t be mistaken for simplicity. The regime continues to present challenges and complexity, especially for those looking to apply the guidance to particular business models and supply chains.

For more information, please contact James Tetley or your usual RSM contact.

Latest related insights