Anthropic Enhances FCA's Supercharged Sandbox with AI Tools

The Financial Conduct Authority (FCA) has announced that Anthropic will support the latest group of firms in its Supercharged Sandbox. This initiative allows companies to safely experiment with advanced AI technologies. Participants will gain access to Anthropic's AI tools, including Claude Code and Claude Cowork, to accelerate development and innovation.

Expanding AI Use Cases

The second cohort in the Supercharged Sandbox will explore various AI applications. These include improving agent-led payments, enhancing fraud detection, strengthening AI governance, and widening access to financial services. The initiative aims to also streamline compliance processes and improve business automation.

Key Participants

Twenty-one organizations, such as Scottish Widows, Money Advice Trust, and TrueLayer, will join this phase. This selection followed a significant surge in applications, increasing by 51% compared to the initial group. This strong interest highlights the industry's need for a safe environment to test AI innovations.

Role of Technology Partners

The Sandbox benefits from ongoing support by NayaOne and NVIDIA. In the first iteration, firms had access to NVIDIA's accelerated computing infrastructure and AI Enterprise software, which will continue for the second group. Jessica Rusu, FCA’s Chief Data Officer, emphasized the demand for trusted experimentation environments, stating that Anthropic's involvement will provide essential technology for innovation.

'The high level of interest in the Supercharged Sandbox demonstrates the demand for trusted environments where firms can experiment safely and responsibly,' said Rusu.

New Training Initiatives

Alongside the Sandbox, the FCA has launched the Agentic Academy, a new specialist AI programme in collaboration with the Centre for Finance, Technology and Entrepreneurship (CFTE). This 10-week course is designed to further support firms in their AI development efforts.

The Road Ahead for AI Experimentation

The FCA's Supercharged Sandbox, launched in 2025, builds on the existing Digital Sandbox by offering enhanced capabilities for AI experimentation. This environment provides firms in the discovery phase with the tools required to innovate responsibly while maintaining the UK's leadership in AI adoption.

The second cohort received 199 applications, a considerable increase from the first cohort's 132. This growing interest reflects the initiative’s significance in the broader push for responsible AI development, as stated in the FCA’s response to the UK Prime Minister's letter on accelerating digital innovation.

Read more about FCA rules related to AI in the regulation section.