The Financial Conduct Authority (FCA) has announced that Barclays, Experian, and UBS are among the latest financial firms to join its live testing initiative for artificial intelligence (AI) applications. The second cohort now includes these major players alongside earlier participants such as Lloyds Banking Group, NatWest, and Monzo. The initiative provides a safe environment for firms to test AI in real-world conditions with appropriate regulatory support and oversight.
The FCA's AI testing initiative is designed to support companies that have advanced significantly in their AI development and are ready to deploy it in live markets. The regulator offers tailored guidance through its technical partner Advai, reflecting a commitment to fostering innovation while managing risks. The FCA stated: “This initiative helps successful applicants explore key questions around risk management and live monitoring to support the responsible deployment of AI for consumers and markets.”
Scope of the second cohort
The second group of participants will test both customer-facing and business-to-business use cases. These include AI-enabled targeted support for investments, credit score insights for consumers, agentic payments, and money laundering detection. According to the regulator, banks are experimenting with a range of technologies, from agentic AI and small language models (SLMs) to emerging innovations such as neurosymbolic AI. Agentic AI refers to systems that can autonomously take actions on behalf of users, while SLMs are smaller, more efficient language models suitable for specific tasks.
The FCA's approach represents a continuation of its broader strategy to engage with financial technology through regulatory sandboxes and innovation hubs. The AI testing initiative, launched in 2025, builds on earlier sandbox programmes that have allowed hundreds of firms to test new products and services under controlled conditions. The regulator plans to publish a report later this year highlighting both good and poor practices observed during these tests.
Regulatory and political context
Separately, UK financial services regulators faced criticism earlier this year from the Treasury Committee of MPs. In a report, MPs accused regulators of taking a “wait-and-see” approach to AI regulation, warning that “the UK public and the country’s finance system are exposed to potential serious harm because regulators in the financial sector are not doing enough.” This criticism came despite the FCA's ongoing initiatives, indicating a gap between regulatory efforts and political expectations.
In response, Sarah Breeden, deputy governor for financial stability at the Bank of England, defended the regulator's actions. She stated: “We share the committee's view that AI has broad, complex and likely long-term implications for how the UK financial system serves the real economy. However, we do not agree with its characterisation that the bank is taking a ‘wait-and-see’ approach to the use of AI in financial services.” Breeden emphasised that the bank has “invested heavily in analysing the current and future risks posed by both the use of AI in financial services, and the broader investment in and adoption of AI across the wider economy.”
The tension between political pressure and regulatory response is not new. In 2023, the UK government published a white paper on AI regulation, advocating a sector-based approach rather than a single central regulator. Since then, the FCA and Bank of England have been developing their own frameworks, including the AI testing initiative and ongoing supervisory work. However, the rapid pace of AI development continues to challenge traditional regulatory timelines.
Recent developments: Anthropic's Mythos
Last week, major UK banks entered discussions with regulators, finance ministries, and national security organisations after the latest Anthropic AI model, named Mythos, unearthed decades-old software vulnerabilities. This event underscored the evolving risk landscape. Meg Hillier, MP and chair of the Treasury Committee, commented: “Recent developments in the world of AI, such as Anthropic’s Project Mythos, show us how fast this transformative technology is moving. It has never been more important that those responsible for maintaining the UK’s financial stability take a proactive approach to understanding and mitigating the risks AI may pose to our financial system.”
The Mythos incident highlighted how AI can retrospectively identify weaknesses in legacy systems, many of which underpin critical financial infrastructure. Banks have relied on decades-old code for core banking operations, and AI-driven vulnerability discovery could force rapid remediation efforts. The discussions between banks and regulators focused on assessing the scale of potential risks and coordinating responses to protect the financial system.
Technologies under the microscope
The AI technologies being tested in the FCA initiative represent a spectrum from mature applications to cutting-edge research. Agentic AI, for example, is attracting attention for its ability to execute multi-step tasks autonomously, such as processing payments or managing customer service interactions. However, it also raises concerns about accountability and error propagation. Small language models (SLMs) offer a more efficient alternative to large models like GPT-4, making them suitable for specific financial tasks where low latency and lower cost are priorities. Neurosymbolic AI, which combines neural networks with symbolic reasoning, is being explored for its potential to provide explainable decisions in credit scoring and fraud detection.
The FCA's technical partner, Advai, specialises in testing AI systems for robustness and safety. Its role includes stress-testing models against adversarial inputs and evaluating their performance in edge cases. This partnership ensures that firms participating in the initiative receive expert guidance on identifying and mitigating risks before full-scale deployment.
Broader implications for financial services
The expansion of the FCA's AI testing initiative reflects a wider trend across global financial centres. In the European Union, the AI Act establishes a risk-based framework that will directly impact financial institutions from 2025. In the United States, the Securities and Exchange Commission has proposed rules requiring companies to disclose AI-related risks. The UK's approach, while differing in its sectoral focus, aims to balance innovation with consumer protection.
For banks and other financial firms, participation in the FCA initiative offers several advantages. It provides a pathway to regulatory approval for AI applications, reduces the uncertainty of compliance, and allows firms to learn from each other's experiences. The FCA has stated that the initiative is not a one-size-fits-all solution but rather a collaborative process tailored to each participant's maturity and use case. This flexibility is crucial given the diversity of AI applications, from chatbots to algorithmic trading.
However, challenges remain. The financial sector is highly regulated, and AI introduces new types of risk, including bias, opacity, and systemic interconnectedness. The FCA's report later this year will likely address these issues, offering insights that could influence future regulation. Additionally, the criticism from the Treasury Committee suggests that MPs expect even faster action. The regulator's response will be watched closely by industry observers and international peers.
Historical context of AI in UK finance
The UK has long been a hub for financial technology innovation. The FCA's regulatory sandbox, launched in 2016, was one of the first of its kind globally. It allowed fintech firms to test products without full regulatory burden. The AI testing initiative extends this concept specifically to artificial intelligence, recognising that AI presents unique challenges not fully addressed by existing sandbox mechanisms. For example, AI systems can evolve over time, adapting to new data, which complicates static testing environments. The live testing approach aims for a new paradigm where monitoring continues after deployment.
Previous cohorts of the AI initiative have included firms using AI for credit risk assessment, fraud detection, and personalised financial advice. The second cohort's focus on agentic payments and money laundering detection indicates a move toward more autonomous and high-risk applications. This progression reflects the industry's growing confidence in AI and the regulator's willingness to engage early with emerging technologies.
The FCA's role extends beyond the testing initiative. It also issues guidance on AI governance, conducts thematic reviews, and collaborates with international regulators through bodies such as the International Organization of Securities Commissions (IOSCO). These efforts contribute to a growing body of knowledge on how to regulate AI in finance, a field that remains in its infancy relative to the pace of technological change.
In summary, the addition of Barclays, Experian, and UBS to the FCA's AI testing initiative marks another step in the UK's approach to regulating AI in financial services. The initiative provides a controlled environment for innovation while gathering data that will inform future policy. Meanwhile, political pressure and recent events like the Anthropic Mythos vulnerability underscore the urgency of proactive regulation. The coming months, with the FCA's report and continued testing, will be critical in shaping how AI is deployed in one of the world's most important financial centres.
Source: ComputerWeekly.com News