Britain’s Financial Conduct Authority is stepping up its efforts to move artificial intelligence testing in financial services beyond theory and into controlled real-world use, as it brings a new group of firms including Barclays, UBS, Experian and Lloyds Banking Group into its AI Live Testing programme.
The FCA, which regulates conduct across the UK’s financial services industry, has been increasingly clear that AI in banking, insurance and markets cannot be judged on model performance alone.
Instead, firms are being pushed to show how AI systems behave once they are deployed into live environments, wrapped in governance, monitoring, human oversight and operational controls.
That shift matters for QA and software testing teams across financial services. As regulators move from consultation and warning into supervised validation, testing is becoming more than a delivery-stage checkpoint.
It is starting to look more like a regulatory artefact in its own right, producing the evidence firms need to show that AI systems are safe, controlled and resilient under real conditions.
Speaking at UK FinTech Week, Jessica Rusu, chief data, information and intelligence officer at the FCA, confirmed that eight new firms have been selected for the initiative, which is being run with technical partner Advai, a London-based specialist in automated AI assurance.
“We’re continuing to collaborate with firms to support the safe and responsible development of AI in UK financial markets,” Rusu said.
“With tailored support from the FCA and Advai, the initiative reflects our commitment to supporting the pace of change in AI, whilst demonstrating how regulators and industry can work together to harness innovation responsibly.”
Big banks enter the next phase
The second wave of firms includes Barclays, Experian, Lloyds Banking Group through Scottish Widows, UBS, Aereve, Coadjute, Go-Cardless and Palindrome.
Their presence underlines how quickly AI testing has moved up the agenda for mainstream financial institutions. This is no longer confined to smaller sandbox participants or early-stage fintech experiments.
The inclusion of major banking and financial names suggests the FCA is drawing increasingly important parts of the sector into a more hands-on model of supervised AI validation.
Applications reflect the fast-evolving shape of AI adoption across financial services, with firms testing a broad mix of technologies and use cases.
According to the FCA, those range from “agentic AI and small language models to emerging solutions such as neurosymbolic AI”, covering both customer-facing and business-to-business deployments.
The use cases themselves point to the widening regulatory perimeter around AI in finance, including “AI-enabled targeted support for investments, credit score insights for consumers, agentic payments, anti-money laundering detection, and Know Your Customer.”
For QA teams, that breadth is significant. It means testing demands are now stretching across consumer outcomes, operational decision-making, fraud controls, payments and compliance workflows.
From pilots to live validation
The programme builds on the FCA’s earlier message that too many firms remain stuck in experimentation mode.
As Ed Towers, head of department in the FCA’s advanced analytics and data science unit, said when the initiative was launched earlier this year: “We’re providing a structured but flexible space where firms can test AI-driven services in real-world conditions, all with our regulatory support and oversight and help from our technical partner, Advai.”
He added that the FCA wants firms to move beyond “POC paralysis”, or what is often described as “perpetual pilots”.
That language has become increasingly important in financial services, where many institutions have explored AI internally but still face challenges around assurance, governance and production readiness.
More importantly for software testing teams, Towers has made clear that the FCA is not looking narrowly at models in isolation.
“We broadly define the AI system as: the actual AI model, information on the deployment context and core risks … governance and human in the loop considerations, evaluation techniques as well as the input and output controls,” he said.
That definition pushes AI assurance well beyond traditional validation exercises. It brings governance, controls and deployment context directly into testing scope, closer to the kinds of full-system checks already expected in cyber resilience and operational risk.
Pressure on banks
The latest intake lands at a moment when UK regulators are sharpening their focus on how AI behaves under stress and in live environments.
The Bank of England has been simulating AI-driven market stress, including possible “herding” behaviour among trading agents, in a sign that AI risk is increasingly being treated as something that must be tested and evidenced, not simply governed on paper.
As Ed Birchall, VP enterprise AI at Nuix, put it: “The Bank of England testing AI-driven systemic risk is a big signal, not just for regulators, but for every financial institution.”
“We’re moving from ‘AI experimentation’ to ‘AI as market infrastructure’,” he argued.
That same shift is visible in the FCA’s own supervisory approach. The regulator has already signalled that testing must extend into real workflows and sensitive environments, including through its planned trial of AI on live financial crime data.
In earlier reporting on that move, Towers said: “Through live testing we want to help UK innovators move safely beyond ‘POC paralysis’, or what is often described as ‘perpetual pilots’.”
For banks and insurers, this means QA teams are being pulled into a broader regulatory agenda, one that links AI validation to digital resilience, third-party oversight and operational accountability.
‘Good and poor practice’ report
The FCA said firms in the latest group began testing in April after applications opened in January 2026. Testing will run until the end of the year, with an evaluation report due in the first quarter of 2027.
The regulator also said it will publish a Good and Poor Practice report for AI in financial services later this year, intended to support firms in the “safe and responsible adoption” of the technology.
That work sits alongside the FCA’s newly published Innovation Insights Report, which found that its Regulatory Sandbox and Innovation Pathways recorded a 49 per cent increase in applications compared with the previous year.
The report also found that FinTech market activity is closely aligned with demand for the FCA’s innovation services, “particularly in fast-growing areas like AI.”
Together, the figures suggest firms are not pulling back from AI experimentation. If anything, they are accelerating.
But the regulatory message is becoming harder to ignore: deployment will increasingly depend on whether firms can produce credible evidence that systems are working as intended in practice.
QA FINANCIAL EVENTS
Why not become a QA Financial subscriber?
It’s entirely FREE
* Receive our weekly newsletter every Wednesday * Get priority invitations to our Forum events *
REGULATION & COMPLIANCE
Looking for more news on regulations and compliance requirements driving developments in software quality engineering at financial firms? Visit our dedicated Regulation & Compliance page here.
READ MORE
- Inside banking’s shift to smarter QA to tackle complexity and risk
- SmartBear CPTO on AI in banking QA: ‘Impressive metrics but no critical scenarios’
- Banks push beyond traditional QA as resilience testing gains ground
- Banking QA professionals warn AI still doesn’t know ‘where the bodies are buried’
- RECAP: The QA Financial Healthcare & Insurance Forum Philadelphia 2026
WATCH NOW
QA FINANCIAL PODCASTS
CLICK HERE TO LISTEN TO OUR EXCLUSIVE CONVERSATIONS
