The Financial Conduct Authority has named the second cohort of firms for its Supercharged Sandbox and said Anthropic will support the group by providing access to Claude, including Claude Code and Claude Cowork. The controlled testing environment is designed to let firms experiment with advanced artificial intelligence safely, and the 21 selected organisations will use it to develop and test applications across payments, fraud prevention, governance, financial inclusion, and compliance automation. The FCA also launched Agentic Academy, a 10-week specialist AI programme for selected firms delivered with the Centre for Finance, Technology and Entrepreneurship. The second cohort includes 21 organisations, among them Scottish Widows, Money Advice Trust and TrueLayer. Their projects will explore use cases such as safer agent-led payments and commerce, stronger fraud and economic crime detection, improved AI governance and accountability, broader access for vulnerable and underserved consumers, and more efficient compliance and business processes. Demand increased from the first round, with 199 applications compared with 132 previously, a 51% rise. The sandbox continues to build on NayaOne infrastructure and NVIDIA support, with second-cohort firms retaining access to accelerated computing infrastructure and NVIDIA AI Enterprise software.