The recent tribunal case involving Get On Board Limited, a company in liquidation by the time of the hearing, highlights critical insights into R&D tax relief claims for AI-driven projects. Get On Board sought R&D tax relief for developing an AI-based system to streamline “Know Your Client” (KYC) procedures in financial services—a notoriously time-consuming process. Since the company was not generating profits, it claimed a surrenderable loss on qualifying R&D expenses, seeking a tax credit.
Background and Dispute
Get On Board’s project aimed to automate KYC tasks through an AI analysis system that handled client risk profiling. The company argued this innovation was groundbreaking, addressing an industry need for more efficient, automated compliance processes. However, HM Revenue & Customs (HMRC) contended that the project did not qualify for R&D relief under the BEIS (Department for Business, Energy, and Industrial Strategy) guidelines. Their primary argument was that Get On Board was not performing unique scientific or technical work, as it used widely available open-source code, AI, and APIs in its solution. According to HMRC, any skilled professional could replicate this approach.
Key Testimonies and Tribunal Considerations
The managing director of Get On Board, a seasoned professional with 25 years of experience in building models for investment banks, provided compelling testimony. He highlighted that the project aimed to create new knowledge and capabilities, leveraging open-source materials in a unique configuration. His R&D report argued that no other system on the market offered this type of automation, suggesting that technical uncertainties existed around whether such a system could even be developed. He further emphasized that he would not have invested years into the project if similar technology were already available.
HMRC’s representative, however, had little technical expertise, which the judge noted as a shortcoming in handling such a case. The judge also remarked that the presence of the company’s software engineer might have clarified key technical questions, although the managing director proved a credible witness.
Tribunal Findings
In its decision, the First-Tier Tribunal recognized several important principles relevant to R&D tax relief in technology:
1. Use of Existing Technologies: The tribunal stated that utilizing open-source code and other existing technologies is common in software development, including AI. This alone does not mean that a project lacks originality or complexity. They noted that if using such resources disqualified projects from R&D relief, few AI projects would qualify, as most involve some level of open-source integration.
2. Advancement and Uncertainty: According to the tribunal, the focus should be on whether the development constitutes a technological advancement, not whether it is entirely novel. This project’s goals involved more than routine adaptation, as it sought to address unique technological uncertainties—challenges that a competent professional could not easily resolve.
3. Qualifying Project Criteria: The tribunal concluded that Get On Board met the criteria for a qualifying R&D project under BEIS guidelines. These included having a clear project with a defined aim, a development process that involved more than copying or minor adaptations, and unresolved technological uncertainties.
Case Management Insights
The judge also commented on the importance of structured evidence in complex R&D cases, suggesting that this case might have been resolved without a tribunal if both parties had presented comprehensive, structured technical arguments upfront. Ideally, the company should have consolidated its technical evidence and arguments in one document, including input from relevant professionals, and HMRC should have responded similarly. Such preparation would have likely streamlined the process and could serve as a best practice for future R&D cases.
Key Takeaways for R&D in AI and Technology
For companies developing AI-driven solutions in compliance-heavy industries, this case provides several insights:
- Evidence is Critical: Consolidating technical arguments and evidence from qualified professionals can strengthen R&D claims and improve case outcomes.
- Understanding BEIS Criteria: Projects using open-source or existing technologies can still qualify for R&D relief if they address significant uncertainties and aim to make genuine technological advancements.
- Preparedness for HMRC Scrutiny: Ensure that R&D projects have thorough documentation to demonstrate unique aspects and technological hurdles. This documentation should align with both the BEIS guidelines and broader Generally Accepted Accounting Principles (GAAP).
This case underscores the need for comprehensive technical documentation and proactive preparation in R&D claims. By taking these steps, companies can better substantiate their R&D projects, potentially avoiding lengthy disputes with HMRC.