FCA opens controlled environment for live AI testing

It will be the first of its kind for UK financial services, with a second round set for April

FCA opens controlled environment for live AI testing

Transformation

By Josh Recamara

The Financial Conduct Authority is working with major financial firms to test artificial intelligence (AI) in a controlled environment to better understand benefits, risks and appropriate safeguards. 

The AI Live Testing initiative will be the first of its kind in UK financial services. The FCA said participants will receive support from the agency's regulatory specialists and technical partner Advai to help assess and implement AI systems safely.

First participants

The first group includes Gain Credit, Homeprotect (Avantia Group), NatWest, Monzo, Santander, Scottish Widows (Lloyds Banking Group) and Snorkl.

Current testing is focused on retail financial services, with use cases being explored for debt resolution support, financial advice, improved customer engagement, complaints handling efficiency and tools that could help consumers make decisions on spending and saving.

Jessica Rusu, chief data, information and intelligence officer at the FCA, said the service is designed to help firms ready for real-world deployment, supporting safe and responsible use of AI in UK markets. Insights generated will also guide the FCA’s future regulatory approach.

Increasing AI adoption

The initiative comes at a time when AI adoption among insurers is accelerating. Surveys suggest that around 95% of UK insurers already use AI for data ingestion, while roughly 60% to 63% have introduced chatbots or generative-AI tools into customer service.

Adoption is also growing in underwriting and claims, with about 43% using AI to support claims handling through automation or predictive analytics. A separate survey found that 89% of insurers are investing in digital and AI-driven transformation, and more than half of board appointments last year were filled by individuals with technology experience - signalling how embedded AI is becoming in insurance strategy and governance.

With AI already touching core processes across the sector, regulatory clarity and safe-testing environments may help insurers scale use cases confidently, particularly where models influence decisions around pricing, claims, fraud detection and customer interaction.

Applications for the second AI Live Testing cohort open in January 2026, with testing to begin in April, the FCA said.

Related Stories

Keep up with the latest news and events

Join our mailing list, it’s free!