FCA: the regulator’s latest sandbox, launched last year (inset: coverage from 9 June 2025), aims to helps firms that are in a ‘discovery and experiment phase’ with AI | Credit: FCA
The UK’s Financial Conduct Authority (FCA) has announced that Anthropic – the prominent San Francisco-headquartered artificial intelligence (AI) company known for developing the Claude large-language models – will support the second group of firms participating in its ‘Supercharged Sandbox’.
The regulator has already been working with another US-headquartered AI technology company – Nvidia – on the AI sandbox (test space) initiative, which the FCA unveiled just over a year ago to help firms ‘experiment safely with AI to support innovation.’
21 organisations (almost entirely companies) in total are participating in the second cohort of the sandbox, which aims to help organisations that are in a ‘discovery and experiment phase’ with AI (Global Government Finance lists the second cohort’s members at the end of this article).
Anthropic will specifically provide the sandbox participants with access to its Claude platform – including Claude Code and Claude Cowork – to help participants ‘speed up’ their development work.
A range of AI uses is being explored in the sandbox, including the testing of solutions designed to: enable safer agent-led payments and commerce; detect fraud and economic crime more effectively; strengthen AI governance and accountability; widen access to financial services for vulnerable and underserved consumers; and streamline compliance and business automation.
RELATED ARTICLE UK’s FCA unveils ‘supercharged sandbox’ to help firms experiment with Nvidia AI – a news story (9 June 2025) on the launch of the sandbox
Anthropic’s benefits to sandbox participants
Applications were received from 199 organisations keen to participate in the cohort – a significant rise from the 132 applications sent in for inclusion in the similarly sized first cohort (which ultimately comprised 11 firms focused on ‘retail innovation and consumer insight’ and 11 firms focused on ‘wholesale and compliance innovation’).
“The high level of interest in the Supercharged Sandbox demonstrates the demand for trusted environments where firms can experiment safely and responsibly,” FCA chief data, intelligence and information officer Jessica Rusu said this week (22 July). “With the support of Anthropic, participants will benefit from the technology they need to accelerate innovation.”
“This is central to our commitment to supporting economic growth – enabling firms to make the most of technological advances while maintaining the UK’s position at the forefront of responsible AI adoption and innovation,” she added.
Anthropic triggered political and financial jitters in April this year when it said that its Claude Mythos tool could outperform humans at some hacking and cybersecurity tasks.
Nvidia itself made global headlines in 2024 when it became the world’s most valuable company by market capitalisation. Participating firms in the supercharged sandbox’s first cohort were given access to accelerated computing infrastructure and Nvidia AI Enterprise software to support the development and testing of their use cases. Participants in the newly announced second cohort will continue to have access to Nvidia’s tools.
The second cohort runs until 31 December 2026. A showcase demonstration day is scheduled for 26 November.
RELATED ARTICLE US forces Anthropic to shut down latest AI models, citing national security concerns – a news story (16 June 2026) from our sister title Global Government Forum
NayaOne’s sandboxing support
The FCA is a big fan of sandboxes, having launched the world’s first regulatory sandbox in 2016. It also operates a digital sandbox and opened a ‘Digital Securities Sandbox’ alongside the Bank of England in September 2024 (see final section of this article for news on this).
Just over three years ago the FCA handed a three-year contract with a value of more than £1m (about $1.26m) to NayaOne, a London-based fintech company, to run its digital sandbox as it put the test-space on a permanent footing. NayaOne has also been providing foundational infrastructure for the supercharged sandbox, with Rusu saying in a speech last year that the authority was “extending” its partnership with the company.
In the FCA’s ‘Strategy 2025-2030’ (published in March last year), the regulator stated that FCA’s innovation services have ‘already helped over 200 financial firms test AI and machine learning technology services.’
The following month (April 2025), it published a proposal for ‘AI live testing’, as part of its existing AI Lab, to support firms ready to use AI. Rusu announced in April that eight firms were participating in the second cohort of companies live testing AI applications. The supercharged sandbox is part of the AI Lab but a different initiative.
Other FCA initiatives on the AI front include a newly announced 10-week specialist programme – named the Agentic Academy – for selected firms. It is delivering this in tandem with the Centre for Finance, Technology and Entrepreneurship (CFTE), a global education company headquartered in London’s Canary Wharf.
RELATED ARTICLE UK’s FCA engages fintech company to run permanent digital sandbox – a news story (28 April 2023) on the FCA handing a three-year contract to NayaOne to run its digital sandbox – see below
HSBC approved to go live in DSS
The DSS has been set up to enable regulators to curate the take-up of digital asset technology in financial markets. It effectively ushers in a ‘test’ regulatory regime for companies to use digital infrastructure such as distributed-ledger technology (DLT) to issue, trade and settle securities in a ‘live’ setting, and will also enable the authorities themselves to hone regulation.
HSBC had been approved by the Bank of England as the first applicant to go live in the DSS, it was announced earlier this month.
HSBC Orion, the London-headquartered global bank’s digital assets platform, will be able to operate within the DSS as a Digital Securities Depository (DSD) for ‘digitally native bond issuance, servicing and settlement, including the DIGIT or digital gilt instrument, and corporate bonds’, the bank announced (15 July).
A gilt is a government liability denominated in sterling and listed on the London Stock Exchange. They have been issued by the UK Debt Management Office (DMO) on behalf of HM Treasury (HMT) since 1998. ‘DIGIT or digital gilt instrument’ is a reference to the government’s planned digital bond issuance, for which HSBC Orion was announced by HMT as platform provider in February.
The now-former chancellor of the exchequer Rachel Reeves announced in her recent Mansion House speech (14 July) that the pilot issuance would take place “by early next year (2027)”.
A four-page ‘Sandbox Approval Notice for Gate 2 of the Digital Securities Sandbox’ for HSBC published on the BoE website the previous day (13 July) states (in respect of activity limits): ‘UK Government debt: To be confirmed; GBP Corporate Bonds: £900m.’
Earlier this month (6 July) the FCA published a 147-page ‘AI and the future of retail financial services’ report (following a review led by executive director Sheldon Mills into how AI could transform retail financial services by 2030 and beyond); and HM Treasury presented the Financial Services AI Adoption Plan (14 July) to encourage safe, rapid scaling and keep the UK competitive globally.
FCA Supercharged Sandbox: second cohort members
Scottish Widows
Money Advice Trust
TrueLayer
calQrisk
Merx Digital Solutions Ltd (SmartDrops)
Aegis Trace
Sardine AI Corp
Zquas
Trustie Labs
Welleness
IntelXview Ltd
RMI Agentic
Ubyx, partnering with Amazon
Deepflow
FSCom
GAI Labs
Condukt
Kaption
Relace
Source: FCA announcement (22 July 2026)