On Monday 9 June, the Financial Conduct Authority (FCA) unveiled its ‘supercharged sandbox’, which will enable banks and fintechs to experiment safely with AI to help drive innovation.
Leveraging AI products from technology development giant NVIDIA, the UK’s financial regulator said the AI sandbox will give UK banks and financial services firms access to better data, technical expertise and regulatory support to speed up innovation.
By enabling firms to access NVIDIA’s full-stack accelerated computing platform to explore AI innovations, the FCA hopes it can support its strategy to drive economic growth by harnessing technological advances like AI.
“The supercharged sandbox significantly upgrades our existing sandbox programme, offering advanced Graphics Processing Unit (GPU) compute resources, enabling users to efficiently train, refine, and experiment with cutting-edge AI models,” explained Jessica Rusu, the FCA’s chief data, intelligence and information officer, at London Tech Week.
The launch comes in the face of 80 per cent of UK banks saying sluggish innovation is hurting their business, fueling talent shortages, resource wastage and eroding their competitive edge, according to a recent research study by core banking fintech SaaScada.
The findings reveal that an inability to experiment and test new solutions or features quickly and cost-effectively is hamstringing banks’ ability to move at pace, with 82 per cent of UK banks stating they need a safe and cost-effective way to experiment and test products.
While the new FCA sandbox looks to fill this gap and bolster innovation, does it go far enough to enact meaningful levels of change, and does it leave anything to be desired? To find out, we explore some views from the industry.
A ‘vital step forward’
For Laurent Descout, CEO and co-founder at investment services company Neo, the newly announced supercharged sandbox is great news for payment firms.
“AI is already transforming the payments landscape, from using machine learning to navigate complex global payment rails and reduce errors, to enhancing the speed and accuracy of transactions.
“But we’re only scratching the surface. AI has the potential to enrich payment messages with real-time insights, forecast cash flows, pinpoint liquidity gaps, optimise reconciliation, and proactively detect fraud. These are game-changing capabilities for treasurers and finance teams.
“The Supercharged Sandbox represents a vital step forward. It will give payment fintechs the environment they need to experiment and scale AI-powered solutions faster. This initiative also has the potential to accelerate innovation, boost economic growth and cement the UK’s leadership position in AI and fintech.”
Retaining a human touch
Suman Rao, MD, UK and Ireland, at wealth management technology company Avaloq, shared her view on the sandbox’s potential impact on wealth managers, and stressed the continuing importance of keeping a human touch despite these technological advancements.
“This announcement from the FCA regarding the launch of a supercharged sandbox to help firms experiment with AI using NVIDIA could be transformative for the wealth management industry, enabling smaller players, who have previously been held back by lack of expertise and access to data, to finally get on board and explore the potential of AI in a supported, low-risk environment,” she explained.
“With our recent research highlighting that 87 per cent of wealth managers across the UK believe AI will be integral to their future work and the UK wealth management as a whole, this initiative will no doubt be a welcome development for the industry.
“With the sandbox promoting AI development in a safe and regulated environment, it will also go some way in providing reassurance to clients, many of whom are currently reluctant to trust AI when it comes to making investment decisions (24 per cent) and financial planning (27 per cent).
“That said, clients value human touch, and it’s unlikely that they will ever feel comfortable delegating decisions completely to AI. With AI development showing no sign of slowing down, the challenge for wealth managers now lies in establishing the correct balance between leveraging the benefits of AI while also maintaining human touch that clients value most.”
Encouraging operational risk elsewhere?
While the FCA plans to drive innovation to boost economic growth with the sandbox, Nikos Maroulis, vice president of AI at Hack The Box, a cybersecurity performance centre, says that, while the programme offers clear benefits to financial firms, firms that operate outside of this regulated environment could feel pressured to keep pace, leading to unplanned negative consequences for the UK’s financial sector.
“The FCA’s sandbox trial empowers financial firms to test AI systems in a controlled environment. For those outside the programme, the pressure to keep pace may lead to less secure forms of experimentation,” explains Maroulis.
“In practice, this is already leading to the growth of shadow AI, where models are developed and deployed without formal oversight, often bypassing internal security reviews. These systems can introduce real operational risk, particularly when cybersecurity teams aren’t involved early to perform offensive security testing against AI-driven applications or develop AI-specific red and blue teaming strategies.
“From a security perspective, this reduces the ability to apply consistent controls or respond effectively during incidents. Teams are left working backwards to secure tools they didn’t assess, stretching already limited resources and increasing burnout.
“As AI adoption accelerates, there’s a need for clear, practical guidance that supports safe implementation across the entire financial sector, not just within selected pilots.
“Governance doesn’t have to slow innovation. When embedded early, it can improve decision-making and reduce uncertainty, giving teams the confidence to experiment responsibly. Sharing learnings from the sandbox, building security capability, and supporting technical upskilling are all central elements towards enabling fast, secure deployment at scale.”
