Aviva's Building a ChatGPT App. Claims Teams Are Holding the Line. Both Are Right.
Why insurance distribution is racing ahead on AI while claims teams hold the line. Where the SOS framework and 3Rs lens say UK service SMEs should actually be deploying AI.
Frequently Asked Questions
Why is insurance distribution adopting AI faster than claims?
Quoting is bounded and rules-based, and a bad quote loses a sale, which is recoverable. Claims decisions touch settlements, vulnerable customers and regulatory duties, so the cost of getting it wrong is asymmetric. That risk profile, not the technology, is what sets the pace.
Where should claims teams actually be using AI right now?
In the admin layer around the decision: FNOL triage, document classification, claims correspondence, TPI rebuttals, basic hire rate challenges, settlement nudges, and fraud signal detection. Decisioning stays with the handler. The AI does the evidence-gathering and drafting around them.
What is the 3Rs framework?
Repetitive, Rules-based and Resource-intensive. A task that scores three out of three is a strong AI candidate. A task that misses on rules (like a complex liability call) should be augmented, not automated. It is the filter we use in every Optimus discovery meeting.
Is the FCA going to regulate AI in UK insurance?
The FCA has confirmed it will regulate AI through its existing fairness, conduct and consumer duty frameworks, rather than building a bespoke AI regime. That puts the burden on firms to evidence human oversight, fair outcomes and a clear audit trail for any AI in the customer journey.
What is one thing to do this month?
Run your top five handler tasks through the 3Rs filter. Pick the highest-scoring admin task, build the smallest possible AI-assisted version with a handler in the loop, and measure the time saved per case. That single number is more useful than any roadmap.
Loading the full page. If it doesn't load, JavaScript is required.