Artificial Intelligence (AI) is becoming increasingly embedded in financial services, from helping firms identify and analyse cyber vulnerabilities to influencing how consumers research investments and make financial decisions.
Two recent publications from the Financial Conduct Authority (FCA) highlight both sides of this development. The FCA’s review of frontier AI and cyber resilience examines how increasingly capable AI models could change the speed and scale at which firms identify cyber vulnerabilities. Separately, new FCA research shows that younger investors are increasingly turning to AI when researching investments.
Together, the findings point to a common challenge: as reliance on AI grows, firms need to ensure that governance, human judgement, operational resilience and consumer protection keep pace.
Frontier AI is Changing the Cyber-Resilience Landscape
The FCA’s multi-firm review explores how firms are using, testing and preparing for frontier AI models in cybersecurity and operational resilience. The regulator describes frontier AI as the most advanced AI models available at any given time and notes that their capabilities can outpace existing regulation and risk-management practices.
These models could help firms identify and analyse cyber vulnerabilities much more quickly. However, the FCA also highlights the potential for the same capabilities to amplify cyber threats if used maliciously.
One of the key findings is that vulnerability discovery may accelerate faster than firms’ ability to respond. AI can increase the speed, scale and volume of vulnerabilities identified, placing additional pressure on validation, remediation, engineering resources and change-management processes.
This means that the technology itself is only part of the equation. The FCA highlights the importance of the environment surrounding an AI model. This includes specialist tooling, robust validation processes, operational guardrails and human expertise. It refers to this environment as “harness engineering”, meaning the controls and processes around an AI model that help make its outputs useful, safe and reliable.
The FCA also stresses that human judgement remains central. While frontier AI can accelerate vulnerability discovery, code analysis and prioritisation, firms still need specialist expertise to validate findings, assess their relevance and make risk-based decisions.
In practice, firms therefore need to consider whether their existing cyber and operational resilience frameworks can cope with a continuous flow of AI-generated findings. The issue is not simply whether a firm can discover more vulnerabilities, but whether it has the capacity to understand, prioritise and remediate them effectively.
Young Investors are Putting their Trust in AI
The importance of human oversight is equally relevant on the consumer side, as AI becomes increasingly influential in investment research.
New FCA research among 18- to 40-year-olds who own or are considering investments found that 56% trust AI tools, compared with 47% for television and radio, 46% for the press and 29% for social media influencers. Four in five less experienced investors have used AI for help with investing, while around two-thirds expect to rely on AI even more over the next year.
However, this growing trust is accompanied by some significant misunderstandings about the protections available to consumers.
The FCA found that 44% of respondents mistakenly believed AI-generated financial information is regulated, while 38% believed it was acceptable to make an investment decision based solely on AI output. Around a third also wrongly believed they would receive compensation from the Financial Services Compensation Scheme (FSCS) or Financial Ombudsman Service if AI advice went wrong.
This is particularly important because general-purpose AI chatbots are not regulated by the FCA. Consumers should therefore not assume that information generated by a general-purpose AI tool carries the same protections as regulated financial advice. Tools specifically designed to provide financial advice may fall within the FCA’s remit, but the distinction is important.
There are nevertheless positive signs. The FCA found that 73% of respondents understood that AI can provide inaccurate information, while 86% recognised the importance of checking the sources referenced by AI.
The message for consumers is therefore not to avoid AI, but to use it carefully: check sources, retain personal judgement and understand the limits of the protection available.
What this Means for Firms?
The two FCA publications point to a common challenge. AI is becoming more capable, more widely used and increasingly trusted, both within financial services firms and by their customers.
For firms, this means considering:
- Governance: Is there clear accountability for AI-related risks and decisions?
- Human oversight: Are AI outputs appropriately reviewed and validated?
- Resilience: Can existing processes cope with the increased speed and volume of AI-generated findings?
- Controls: Are appropriate guardrails, access controls and escalation processes in place?
- Consumer protection: Are customers aware of the limitations of AI-generated financial information?
These considerations should be viewed in the context of firms’ existing regulatory obligations. The FCA’s frontier AI multi-firm review does not introduce new rules, guidance or regulatory expectations. Instead, its findings should be considered alongside applicable existing requirements, including the FCA’s operational resilience framework and, where relevant, the Consumer Duty.
AI presents significant opportunities for financial services, but those opportunities need to be matched by robust governance and resilience.
As AI continues to develop, the question is no longer whether it will become part of financial services, but whether firms are ready to govern it responsibly.
Book a Meeting with a Complyport SME
To discuss how developments in AI may affect your firm’s governance, operational resilience and regulatory compliance arrangements, contact us today to book a meeting with a Complyport Subject Matter Expert.
Ask ViCA, your Virtual Compliance Assistant.
Access instant answers on regulatory changes.
Claim your complimentary 20 queries today! Register here: https://vica.chat
CPT social media:
#AI #ArtificialIntelligence #FinancialServices #CyberResilience #OperationalResilience #Compliance #FCA #FinTech #AIRegulation #Investment
