Edinburgh-headquartered Scottish Widows is one of 21 firms selected for the second cohort of the Financial Conduct Authority’s Supercharged Sandbox, the UK regulator’s AI innovation programme, which is now being backed by Anthropic.
Anthropic will give participating firms access to its Claude AI models, alongside its Claude Code and Claude Cowork tools. The sandbox already runs on NayaOne and NVIDIA infrastructure. The cohort was selected from 199 applications — a 51% increase on the inaugural intake.
Alongside Scottish Widows, the named participants include payments infrastructure firm TrueLayer and debt-advice charity Money Advice Trust. Use cases set out by the FCA span fraud detection, financial crime prevention, AI governance, risk management, agent-led payments and commerce, compliance automation, and expanding access to financial services for underserved and vulnerable consumers.
According to the FCA, “the programme aims to support innovation while ensuring new AI applications are developed in a safe, accountable, and regulated environment.”
The expansion sits within the FCA’s broader strategy of promoting responsible AI adoption without introducing AI-specific regulation, and is complemented by a 10-week Agentic Academy delivered with the Centre for Finance, Technology and Entrepreneurship. For Scotland, it puts one of the country’s largest financial-services employers directly into a UK regulator-backed environment for testing frontier-AI use cases.
