Wednesday, 24 June 2026
The UK’s financial regulator is reassessing its approach to oversight as artificial intelligence becomes increasingly embedded across financial services, according to a speech delivered by Nikhil Rathi at techUK’s Agents of Change conference.
Rathi, chief executive of the Financial Conduct Authority (FCA), said financial services will play a central role in the UK’s ambitions to become a leading AI economy by providing the capital, infrastructure and trust needed to support wider adoption of AI technologies.
According to the FCA, more than 80 percent of financial services firms are already using AI in some form. The focus is now shifting from experimentation to scaling deployment across retail and wholesale markets.
One area highlighted by the regulator is the emergence of so-called agentic AI systems, which are capable of carrying out actions and transactions rather than simply generating content or analysing data. Rathi said such systems could support functions ranging from personal financial management and investment strategies to liquidity management and trading operations.
He stressed, however, that accountability for regulated activities must remain clearly assigned and subject to appropriate human oversight.
The FCA also identified tokenisation as a significant development for financial markets. The regulator recently approved the launch of the UK’s first natively tokenised authorised investment fund by Baillie Gifford and Bank of New York Mellon. The FCA and the Bank of England are also consulting on the future development of tokenised wholesale markets.
Rathi argued that the pace of technological change is challenging traditional regulatory approaches and that legislation alone cannot keep pace with developments in AI. As a result, the FCA is placing greater emphasis on risk-based supervision, market stewardship and collaboration with industry participants.
The regulator is also exploring the use of agentic AI internally to enhance market surveillance and detect potential market abuse more quickly. The FCA currently processes around one billion rows of data daily as part of its supervisory activities.
The speech highlighted the growing importance of competition policy as AI lowers barriers to entry and accelerates market change. Rathi said the FCA expects to make more frequent use of its system-wide powers under UK competition legislation to address emerging risks and market developments.
At the same time, the regulator acknowledged that collaboration between firms may become increasingly necessary in areas such as open finance, data sharing and AI development.
The FCA also warned of growing resilience and cybersecurity risks linked to increasing dependence on cloud infrastructure, AI model providers and external technology suppliers. According to industry data cited by the regulator, the UK lost nearly £1.3 billion to payment fraud in the past year, with most authorised fraud cases originating through social media and messaging platforms.
Rathi noted that 98 percent of operational incidents reported to the FCA last year were related to technology or cyber issues. He said financial institutions should strengthen governance around third-party dependencies and ensure boards understand the risks associated with AI adoption.
To support innovation, the FCA continues to expand initiatives including its Supercharged Sandbox testing environment, AI Lab and AI Consortium, which it operates together with the Bank of England. The regulator plans to publish further guidance on AI practices later this year, alongside a review examining how AI could reshape retail financial services.
The FCA said its objective is to support the safe and responsible adoption of AI while maintaining competition, resilience and consumer confidence across the financial system.
Source: FCA
