Anthropic joins UK FCA AI regulatory sandbox as second cohort opens

Anthropic is joining the UK Financial Conduct Authority’s AI regulatory sandbox as the second cohort of the FCA’s Supercharged Sandbox launches. Under the program, Anthropic will provide its Claude AI models to companies participating in the next round of testing focused on financial services use cases.

The development matters because it gives regulated firms a structured environment to experiment with AI tools while the FCA observes how those systems perform in practice. For financial companies, the sandbox can help explore AI applications in compliance, operations, and customer-facing services under regulator oversight.

The FCA’s Supercharged Sandbox is designed to support testing of emerging technologies in financial markets, with a focus on responsible deployment. Anthropic’s participation adds a major AI provider to the initiative and expands the range of tools available to the companies involved.

For readers following crypto and broader financial technology trends, the move is another sign that regulators are trying to understand and shape AI adoption rather than simply react to it later. While the sandbox is aimed at financial services more broadly, the outcomes could inform how AI tools are assessed across adjacent sectors that rely on automation and data-heavy decision-making.

The second cohort’s launch shows the FCA continuing to build out its framework for supervised experimentation. Anthropic’s role will be to support participating firms with access to Claude as they test potential applications within the sandbox.