The UK Financial Conduct Authority (FCA) has announced a collaboration with leading artificial intelligence firm Anthropic to enhance its “Supercharged Sandbox” — a regulatory testing environment where firms can experiment with AI products in financial services. The partnership will give accepted participants access to Claude Code and Claude Cowork, Anthropic’s advanced AI tools.
The sandbox, already backed by NVIDIA for accelerated computing infrastructure and NayaOne as a platform partner, is designed to spur innovation while ensuring compliance. Twenty-one firms have been selected for the second cohort, including financial institutions like Scottish Widows, fraud-prevention startups such as Sardine AI Corp, and payment solutions provider TrueLayer. Participants span areas like AI governance, consumer protection, and business automation.
Key objectives outlined by the FCA include safer agent-led payments, more effective detection of fraud and economic crime, stronger AI accountability, wider access to financial services for vulnerable consumers, and streamlined compliance processes. Jessica Rusu, the FCA’s Chief Data, Intelligence and Information Officer, emphasized that the Anthropic partnership will “accelerate innovation.” The regulator also launched an Agentic Academy, a 10-week specialist AI program for approved firms, reflecting the growing adoption of AI — which is already used by an estimated 75% of UK financial services firms.
