Key Facts
- •Get Onbord Limited (GOL), in liquidation, appealed HMRC's rejection of its R&D tax credit claim under section 1054 of the Corporation Tax Act 2009.
- •GOL's project involved developing a novel, automated AI analysis process for KYC verification and risk profiling.
- •HMRC rejected the claim, arguing the project did not advance overall knowledge or capability.
- •A procedural issue arose concerning GOL's representation after entering liquidation.
- •The Tribunal considered whether GOL's project met the 'research and development' tests in the BEIS Guidelines.
Legal Principles
For a claim under section 1054 CTA 2009, expenditure must be on 'research and development' as defined by section 1138 CTA 2010, incorporating the BEIS Guidelines.
Corporation Tax Act 2009 (CTA 2009), Corporation Tax Act 2010 (CTA 2010), Income Tax Act 2007 (ITA 2007), Research and Development (Prescribed Activities) Regulations 2004, BEIS Guidelines.
The BEIS Guidelines require a strict interpretation; purposive construction is not appropriate.
Gripple v HMRC [2010] EWHC 1609 (Ch), Hadee v HMRC [2020] UKFTT 0497 (TC), Flame Tree Publishing v HMRC [2024] UKFTT 00349 (TC).
The burden of proof is on the claimant (GOL) to substantiate its entitlement to the R&D tax credit, but the evidential burden may shift.
Brady v Group Lotus plc [1987] STC 635, Wood v Holden [2006] EWCA Civ 26.
R&D claims require substantial evidence, including from a 'competent professional', to prove the project advanced overall scientific or technological knowledge, not just internal company knowledge.
AHK Recruitment Limited v HMRC [2020] UKFTT 232 (TC), Flame Tree Publishing Limited v HMRC [2024] UKFTT 00349 (TC).
Outcomes
Appeal allowed; GOL's project constituted research and development.
The Tribunal found GOL's project sought to resolve technological uncertainties by developing an AI system for KYC verification and risk profiling, a capability not readily available or deducible by a competent professional. The project involved significant new code and addressed a real-world problem in a novel way. While existing technologies were used, the overall advance was deemed non-routine.