FCA's AI Push: Anthropic Joins Supercharged Sandbox with Claude Models

The UK Financial Conduct Authority (FCA) has launched the second cohort of its Supercharged Sandbox program, designed to foster the safe and responsible development of artificial intelligence (AI) applications within the financial services sector. In a significant move, leading AI company Anthropic will support this initiative by making its Claude AI models available to participating firms, aiming to accelerate financial innovation.

The second cohort of the FCA's Supercharged Sandbox program commenced on July 13, 2026, and is scheduled to run until December 31, 2026. Twenty-one organizations were selected from 199 applications, representing a 51% increase compared to the first cohort, underscoring the growing demand for regulated environments to experiment with AI systems. Participants will gain access to Anthropic's Claude AI models, including Claude Code and Claude Cowork. These models will be utilized to test a range of use cases, such as enabling safer agent-led payments, enhancing fraud detection, strengthening AI governance and accountability, widening access to financial services for vulnerable and underserved consumers, and streamlining compliance and business automation. Prominent organizations like Scottish Widows, Money Advice Trust, and TrueLayer are among the participants in this second group.

This program builds upon the FCA's existing partnerships with NVIDIA, which provides GPU infrastructure and AI Enterprise software, and NayaOne, offering digital sandbox infrastructure. Furthermore, the FCA has introduced a 10-week Agentic Academy in collaboration with the Centre for Finance, Technology and Entrepreneurship (CFTE) to assist firms in developing and deploying AI agents within financial services. The Supercharged Sandbox is specifically tailored for early-stage AI projects focusing on discovery and experimentation, such as proofs of concept and prototypes.

This development is expected to have a significant impact on promoting safe and responsible innovation within the UK financial services sector. The program helps firms accelerate their journey from concept to deployment while reducing the cost of AI experimentation, thereby making innovation more accessible to a broader range of companies. Jessica Rusu, FCA's Chief Data, Intelligence, and Information Officer, emphasized that supporting economic growth and maintaining the UK's leadership in responsible AI adoption and innovation are central to such initiatives.

The FCA's strategy focuses on leveraging existing regulatory frameworks rather than introducing additional regulations specifically for AI. The Supercharged Sandbox is an integral part of the FCA AI Lab's next-generation experimentation platform and complements the upcoming AI Live Testing service, which supports firms transitioning from sandbox experimentation to live deployment. This initiative positions the UK at the forefront of responsible AI innovation in finance.

The high level of interest in the program, evidenced by the significant increase in applications, demonstrates a strong market demand for trusted environments for AI experimentation. Analysts suggest that such initiatives could bridge the gap between AI's potential and the stringent safety requirements of financial services, potentially setting a global blueprint for AI regulatory sandboxes. The insights garnered from the program are expected to inform future regulatory thinking.