The government has published its response to the consultation 'Research and development tax relief clearances'. The consultation highlighted the need for reform of the current advance assurance system.
In March 2025, the government published its consultation, 'Research and development on advance clearance procedure', seeking opinions on widening the use of clearances to avoid error and fraud.
The government was looking at options to reduce error and fraud, which should provide certainty to businesses and improve the taxpayer experience.
The government has highlighted its commitment to maintaining the current rates in both the merged R&D Expenditure Credit (RDEC) scheme and the Enhanced R&D Intensive Support (ERIS) for Small and Medium-sized Enterprises.
- This means companies continue to receive between £15 and £17 for every £100 they spend on R&D.
The government recognises that R&D plays a part in driving innovation and economic growth, and how it may benefit society; however, in recent years, non-compliance has reached unacceptable levels. The government aims to tackle this non-compliance and ensure a fairer system for everyone.
The consultation explored the following areas:
- The current Advance Assurance System; if it is still effective, and can improvements be made?
- How a new future clearance system might look, including eligibility criteria, delivery and timing and whether this should be voluntary or mandatory.
Improving the current Advance Assurance Scheme (AAS)
The current AAS has seen a low and declining uptake. The respondents to the consultation suggested this may be due to:
- The eligibility criteria are too restrictive, with only SME claimants eligible.
- A lack of awareness around the system, particularly among agents and first-time claimants.
- Administrative burdens, including overly long processing time, unclear guidance and inconsistent feedback.
Overall, the respondents were in support of a reform of the current AAS, with the following factors being highlighted for consideration:
Reducing the compliance burden.
- Provide early certainty.
- Improve decision-making.
- Flexibility and accessibility.
- Clearer guidance and improved communication.
Voluntary Mandatory Assurance
A large proportion of the respondents indicated a preference for a voluntary model over a mandatory model.
- It was suggested that a voluntary model would be flexible and more suitable for businesses with varying needs.
- It would allow companies to seek certainty where required.
- It would not impose additional requirements on those who are confident in making claims.
- It was thought that a voluntary model would be particularly helpful for SME's who are often unable to manage complex compliance processes.
The support for a mandatory model was more cautious:
- Respondents favoured a targeted, risk-based approach.
- It was suggested that a mandatory model may be appropriate for first-time claimants and sectors where non-compliance is high.
- There were concerns with increased costs, administrative complexity and the potential to deter claims.
Preferred Assurance timing and focus
It was highlighted among respondents that pre-claim assurance:
- Option B was seen as the most practical stage to offer certainty, with most businesses generally having a clearer understanding of their R&D activities at that point.
- Option A, pre-activity, was seen as more relevant for large and complex R&D projects.
- Option C, post-claim assurance, was considered less useful overall.
Minimum Expenditure Threshold
It was suggested that a threshold could help HMRC focus resources on higher-value claims, reduce administrative burden and deter fraud. Thresholds between £10,000 and £25,000 were suggested.
However, many respondents noted that impactful R&D is not always linked to high expenditure.
- A threshold could exclude innovative firms from accessing support.
- Discourage legitimate claims.
Improving HMRC capability and communication
It was highlighted by respondents that:
- HMRC need to strengthen their capability and engagement with claimants.
- Caseworkers lacked sector expertise.
- Better training and specialist knowledge are required to improve consistency and confidence in decision-making.
Next steps
HMRC will launch a limited pilot of a new targeted Advance Assurance Service. The existing advanced assurance will continue to run while the pilot runs.
Further detail is expected in due course.
Current Advance Assurance Scheme
Only a small proportion of respondents had used the current AAS. Those who had used it felt that:
- It benefited from early certainty and reduced the risk of enquiry.
- The eligibility criteria are too restrictive.
- The process was not flexible.
- Turnaround times were too long.
Respondents raised concerns about inconsistent decision-making and vague feedback, with some respondents suggesting that the scheme was an invitation for scrutiny rather than a source of reassurance.
Many respondents expressed the requirement for a new, more flexible and accessible service.
Preferred Assurance Timing and Focus
Option B, the pre-claim assurance model, was the most popular option, being seen as practical and beneficial whilst offering certainty at a point where the project details are clear but before the actual claim is submitted.
Option A, pre-activity, had a mixed response:
- Larger businesses or those who undertake more complex claims saw value in early engagement.
- It was felt that this option would be of limited use to SME's.
- Concern was also raised about HMRC not having the capacity to deliver timely and meaningful support at this earlier stage.
Option C, post-claim assurance, was seen as the least useful option and had limited support.
Voluntary v Mandatory
Mandatory assurance was not favoured and typically respondents felt it would increase costs and administrative complexity as well as discourage legitimate claims.
To mitigate these risks, respondents emphasised the importance of transparency in how mandatory assurance might be applied. There would need to be clear criteria for identifying high-risk cases and procedural safeguards.
Voluntary assurance was generally favoured among respondents, with many highlighting the flexibility and suitability for businesses with different levels of complexity in their cases.
- It was seen as a way to support businesses that seek further clarity when claiming.
- It was particularly welcomed by SME's.
There was a strong opposition to introducing paid assurance services, with suggestions that this would discourage smaller businesses and reduce access to support.
Eligibility and targeting
Many respondents were against targeting a voluntary assurance service across certain sectors or to high-growth companies, highlighting that innovation occurs across all industries.
- Some acknowledged that prioritising strategic sectors may align with wider government objectives.
- There was concern about how these categories would be defined.
- Respondents favoured a more broadly accessible service which was open to all.
- There was support for certain groups benefitting from prioritised support.
- Any targeting should be based on clear and objective criteria, such as the complexity of the claim or the experience of the claimant.
Minimum Expenditure Thresholds
There were mixed views on the threshold:
- It was thought there was some benefit to a threshold, particularly in assisting HMRC to focus resources on higher claims.
- HMRC data indicate that there are higher levels of non-compliance in claims with qualifying expenditure of less than £50,000, which was a factor in determining the level the threshold should be introduced.
- Supporters of the threshold suggested it could streamline compliance, improving the integrity of the R&D system overall.
- Many respondents believed there would be a broader impact of introducing a threshold.
- A threshold may exclude smaller businesses with lower budgets.
- Excluding smaller businesses may stifle innovation.
- A threshold may also encourage artificial inflation of claims.
Role of agents
It was highlighted that agents play an important role in the process of R&D claims and that any future model must recognise this.
- Many claimants rely on agents for guidance, preparing claims and corresponding with HMRC.
- Suggestions were made that agents must be allowed to submit and manage clearance applications.
- There must be appropriate authority and visibility built into the system from he outset.
Many respondents highlighted the lack of formal qualification or regulatory requirements for making claims, increasing errors and fraud.
- Agent accreditation was suggested to deter fraudulent companies from making claims.
- More enforcement of the code of conduct.
- Some respondents suggested publishing a list of approved agents alongside a register of agents with a history of non-compliance.
- Other suggestions included:
- Online portals.
- Agent-led submissions.
- Feedback forums.
Other suggestions
These included:
- Making guidance clearer and providing specific sector examples.
- Better training for HMRC staff with the use of external experts.
- Improved communication with HMRC, such as helplines and webchat.
- A digital claims tracker or portal.
- More thorough checks on first-time claims.
- Using AI and data analytics to support risk profiling.
Useful guides on this topic
R&D: Advance Assurance
What is Research and Development (R&D) Advance Assurance? How can I apply? What are the conditions? What are the benefits?
Research & Development Tax Reliefs
What is R&D Relief? How does it work? Why does the size of the company matter? What is sub-contracted R&D? How do I write an R&D Report?
R&D: Additional Information Form (AIF)
What is an Additional Information Form (AIF)? When is a company required to submit an AIF? What information should be included in an AIF? How do I submit an AIF?
R&D: Claim Notification
What is a claim notification? When does a company need to submit a claim notification? What is the deadline? What happens if a company fails to submit a claim notification? What information should be included on a claim notification?
External link
Research and development tax relief advance clearances - summary of responses