Ministers have been urged to strengthen regulators’ powers to protect consumers against the potential risks of AI, following a recent landmark review by the Financial Conduct Authority (FCA).
The review, which examined how AI is expected to reshape financial services from 2030 onwards, found that firms are already moving away from traditional human-led processes towards AI-enabled services that will increasingly influence everyday consumer decisions.
This comes following new research from Zable, which reveals 1 in 10 Brits are already using AI platforms for financial planning. The findings below highlight key insights from the survey and reinforce recent FCA warnings around the risks of relying on unregulated AI-generated financial advice.
Key findings from the survey:
- Gen Z (adults aged 18–24) are the least likely to verify financial advice from AI tools and social media platforms. Just under 80% admit they do not feel confident when it comes to verifying financial advice before acting on it.
- When it comes to monthly financial planning, Glasgow residents are the most likely to seek advice from AI, with 15% saying they use AI tools. London and Bristol follow closely behind at 14%, while Brighton ranks next at 13%.
- 40% of millennials use AI tools for monthly budgeting and financial planning advice, compared with just under a third of Gen Z and 28% of credit card owners aged 35 to 44 turning to AI for these areas.
Alongside the survey, Zable tested popular AI platforms to highlight the risks of relying on Artificial intelligence for financial guidance. The results revealed widespread inaccuracies, with many platforms providing US-focused recommendations such as 401(k) pensions, FDIC insurance and other products irrelevant to UK consumers.
Even the best-performing model, Claude, passed only six of the nine tests and still returned some outdated UK information. Grok failed all nine questions, while ChatGPT passed just two and Gemini three, reinforcing the importance of seeking regulated financial advice rather than relying on AI-generated responses.
