The Model Office Blog

The FCA Advice Boundary and AI

[fa icon="calendar"] Apr 23, 2024 4:04:50 PM / by Chris Davies

With the FCA stipulating that advice boundary review (along with the Consumer Duty) is its top priority for 2024, firm’s should now engage how tech can enable streamlined advice and support the mass unadvised.  The FCA Discussion Paper 23/5 delves into the nuanced boundary between financial advice and guidance, raising pertinent questions about the application of artificial intelligence (AI) in navigating this complex terrain.

Given the failure of the FCA CP22/2 with its much-criticised core investment advice strategy, (an ISA selling machine) the regulator is acknowledging their needs to be a distinct move on their part to relaxing some of the advice regulations and have effectively come up with a ‘focused approach’ given the rise in fintech and AI to part-enable this.

At its core, this revised  initiative seeks to refine the definitions and regulatory frameworks surrounding financial advice and guidance. It acknowledges the growing prevalence of digital tools and AI-driven solutions in delivering financial services, prompting a reassessment of traditional distinctions between advice and guidance.

One of the primary challenges in this discussion is delineating the level of human involvement necessary for a service to be classified as financial advice. Historically, financial advice has been characterised by personalised recommendations tailored to an individual's specific circumstances, whereas guidance offers more general information and educational support without personalised recommendations.

The FCA have settled on a three-tier strategy across;

  • Clarifying the boundary:
    • Using non-Handbook and Perimeter guidance to set out scenarios of consumer support that would not constitute a personal recommendation.
    • Simplifying existing guidance to give greater certainty to firms that they are not giving a personal recommendation.
  • Targeted support to segmented clients: delivery of information and clear communications on products and services based on similar client needs
    • Use limited personal information about a customer and their circumstances to provide support to consumers to help them make an informed decision
    • Offer targeted support without explicit charges (i.e. without upfront fees specifically and exclusively relating to the service provision of targeted support), but with a clear disclosure of how a consumer is paying for the service through other associated charges.
  • Simplified advice focused suitability on specific requirements
    • Simplified advice would result in a recommendation that is personalised to an individual consumer’s circumstances – i.e. ‘you specifically’ would benefit from this action. 
    • Simplified advice is a one-off service whereby the firm would take into account only relevant information about a specific consumer need

AI presents both opportunities and challenges in this context. On one hand, AI-driven algorithms can analyse vast amounts of data to provide tailored insights and recommendations, blurring the line between guidance and advice. On the other hand, AI can also enhance the delivery of guidance by offering personalised, yet non-binding, suggestions based on user preferences and behaviour.

So, how can AI contribute to navigating the financial advice-guidance boundary in a manner that prioritizes consumer protection and regulatory compliance?

1. Data driven tools: AI operates by using statistics and probability. This means, as long as a firm’s data is high quality (we know) then an AI assessment tool can work through the probabilities and likelihood for products meeting focused mass market client needs. By incorporating this approach, the AI could streamline fact finding, risk assessment, cash flow and research processes to deliver suitable guidance incorporating the FCA ‘people like you’ strategy and reduce time and costs of delivery along the way.

2. Compliance, Transparency and Disclosure: AI-powered financial tools should be transparent about their capabilities and limitations. A well-trained AI could then assess guidance meets FCA requirements and flag if it strays into advice. Clear disclosure of whether a service is providing guidance or advice, along with the level of human involvement in decision-making, is essential for informed consumer choices.

3. Algorithmic Accountability: Regulators need to ensure that AI algorithms used in financial services are fair, transparent, and accountable. Robust oversight mechanisms can help mitigate the risks of algorithmic bias or opacity, ensuring that AI-driven recommendations are aligned with regulatory standards and consumer interests.

4. Education and Empowerment: AI can be leveraged to enhance financial literacy and empower consumers to make informed decisions. By providing tailored, educational content, interactive simulations, and personalised feedback, AI-driven guidance tools can help individuals better understand their financial options and goals.

5. Dynamic Regulatory Frameworks: Regulatory frameworks must be adaptable to accommodate technological advancements and evolving consumer preferences. Continuous dialogue between regulators, industry stakeholders, and consumer advocacy groups is essential for developing agile regulatory frameworks that balance innovation with consumer protection.

Ultimately, FCA’s Discussion Paper 23/5 underscores the need for a nuanced approach to regulating AI-driven financial services, particularly in the context of the advice-guidance boundary. By fostering collaboration and innovation while upholding consumer safeguards, regulators and retail intermediary firms can harness the potential of AI to enhance consumer financial well-being and promote a more inclusive and transparent financial ecosystem.

Please click the below icon to learn more about MO RegTech today..

New Call-to-action


Topics: Financial regulation, client engagement, regtech, Risk management, practice management, FCA, advice, compliance, consumer duty, boundary

Chris Davies

Written by Chris Davies

Subscribe to Email Updates

Recent Posts