FCA/BoE/PRA tighten scrutiny of insurers' AI/ML and Consumer Duty enforcement
Key Questions
What is the FCA's approach to regulating AI use by insurers?
The FCA is not introducing new AI-specific rules but is intensifying scrutiny of how 750 firms govern the technology, with a focus on explainability, auditable models, and human-in-the-loop requirements targeted for 2026.
What does the Guidewire report reveal about UK pet owners' views on AI in insurance?
The report shows that only 28% of UK pet owners would trust an AI scan for insurance decisions, while 50% would not, highlighting significant consumer skepticism amid broader concerns over AI denials and Consumer Duty compliance.
What AI-related risks and governance issues are highlighted for insurers?
Key concerns include a 47% risk of decisions based on hallucinated AI content, vendor and privacy flags from the Palantir trial, actuarial skill gaps, and the need for a five-tier maturity model, alongside BoE/PRA model risk principles due in April 2026.
FCA is not issuing new AI rules but actively scrutinising how 750 firms govern the technology, with plans to publish good/bad practices later in 2026. Intensifying explainability, auditable models, human-in-loop. Palantir trial raises vendor/privacy flags. BoE/PRA roadmap (Apr 2026) adds model risk principles. AI hallucination risk (47% decisions based on hallucinated content) flagged as emerging governance blind spot. AI Risk Economy video adds actuarial gap and five-tier maturity model. New case study shows FCA calling firms 3 months post self-attestation for fair value/vulnerable customer MI gaps. Aviva's record £230m fraud detection using AI underscores escalating fraud sophistication. Guidewire report shows 28% of UK pet owners would trust an AI scan, 50% wouldn't. FCA fines BancTrust CEO £99,600 for non-disclosure under SMCR.