Digitisation in insurance: what’s changing and where R&D sits

Why digitisation in insurance looks different now

A few years ago, “digital transformation” often meant moving legacy journeys online. Now the conversation is about building decision-making engines that can work at scale across underwriting, claims and service, while staying explainable and compliant.  

Two forces are doing most of the heavy lifting:

1) AI is moving from pilots to production

Many insurers have experimented with AI, but the current push is about scaling it across core domains, not keeping it in innovation labs. Claims automation, document understanding, triage and underwriting support are typical starting points because they are high volume and data rich.  

2) Consumer Duty is raising the bar on digital journeys

In the UK, the FCA has been clear that digital journeys must support good customer outcomes. Poor design can nudge customers into the wrong decisions or hide key information. That makes journey design, testing and monitoring a board-level issue, not just UX polish.  

What’s changing in practice

Claims: faster decisions with better evidence

Digitised claims is moving towards straight-through processing for simpler cases and better assisted handling for complex ones. This includes automated document capture, faster liability triage, and damage assessment workflows, plus improved customer updates that reduce inbound calls.  

Underwriting: better risk signals, not just faster quotes

Underwriting digitisation is shifting from basic rule engines to richer models that use more real-time signals. The challenge is balancing speed with governance, fairness, and clear justification for decisions.

Servicing: less admin, more useful support

A lot of value sits in the unglamorous bits: policy changes, renewals, mid-term adjustments, complaints and customer comms. Automating these safely is often where insurers see the biggest operational wins because the volumes are high and the processes are repetitive.  

Where R&D can show up in insurance digitisation

Not every IT project is R&D for tax purposes. The R&D opportunities tend to appear when you are solving technical uncertainties that a skilled professional could not easily resolve using existing knowledge.

Examples that can often involve genuine R&D work may include:

• Building new methods to detect fraud in low-signal environments without creating unfair customer outcomes
• Creating explainable underwriting models that meet performance targets while staying auditable and controllable
• Developing damage assessment or claims triage workflows that work reliably across many edge cases and data sources
• Using privacy-preserving analytics to improve pricing or risk selection without exposing sensitive data
• Designing multi-model pipelines (rules, ML, language models) that can operate safely at scale with measurable controls  

If your work includes testing, experimentation, and iteration to overcome scientific or technological uncertainty, it is worth reviewing whether it qualifies.

How the merged R&D tax relief schemes can support insurance innovation

For accounting periods beginning on or after 1 April 2024, the previous SME and RDEC schemes no longer apply. Most companies now claim through the merged R&D expenditure credit scheme, with ERIS available for eligible loss-making R&D-intensive SMEs.  

Merged R&D tax scheme

Most companies claim an above-the-line R&D expenditure credit on qualifying spend. The headline rate is 20%, and the credit is taxable, so the net benefit depends on your Corporation Tax position.  

ERIS

ERIS is only for loss-making SMEs that are R&D-intensive. The intensity threshold is broadly 30% of total relevant expenditure, and the payable credit can be worth 14.5% of the surrenderable loss, subject to the rules and caps.  

A key watch-out for insurance teams using global delivery

Rules restricting certain overseas subcontractor and overseas worker costs apply, with limited exceptions. If your data engineering, modelling or build work is delivered outside the UK, it is worth checking eligibility early so you do not model a benefit you cannot claim.  

What HMRC will expect to see

If you want a claim that stands up to scrutiny, focus on clarity and traceability:

1) A tight technical narrative

Spell out the uncertainty, what you tried, what changed, and what you learned. Keep it technical, not marketing.

2) A clean link between activity and cost

Show who did the R&D, what they did, and how you apportioned time and costs.

3) Do not miss the admin steps and deadlines

Claims can be blocked if process steps are missed, including claim notification where required and the Additional Information Form (AIF). Deadlines also matter so you don't miss the opportunity even when you know the work qualifies.

Use our calculator for a quick estimate

If you want a quick indicative figure, try our R&D tax calculator. It provides an illustrative estimate only and is not a substitute for advice. Your actual outcome depends on your qualifying expenditure, your wider tax position, and scheme rules such as caps and restrictions.