Copy article

UK Treasury orders AI skills blueprint for financial services. But can consumers wait until 2027?

ended 07. November 2025

HM Treasury has formally commissioned the Financial Services Skills Commission to map the skills, training and innovation the sector needs for safe, productive AI, working with the City of London Corporation, TheCityUK, Lloyds Banking Group and PwC. The final report is due by mid-2027, with scope to deliver earlier or in stages.

AI isn’t a distant prospect in money management: it’s already the UK’s No.1 use case. Lloyds’ new Consumer Digital Index finds 28m adults (over half of us) used AI in the past year to help manage their money. Trust and safety are now the battlegrounds: people want accuracy, security and advice that’s genuinely tailored to them. If skills, governance and product standards don’t keep pace, we risk normalising unregulated “AI money advice” while regulated firms scramble to retrofit controls.

This commission is an opportunity to set practical guardrails: define workforce capabilities beyond a tiny pool of specialists, establish transparent human-in-the-loop practices, and give firms a route to deploy AI that boosts productivity without outsourcing judgement to black boxes. Existing industry work on AI skills frameworks and data access shows there’s a base to build on; the question is whether government, industry and educators can turn pilots into sector-wide competence fast enough to protect consumers and competitiveness.

We want your views:

  • If more than half the public already lean on AI for money decisions, what “minimum safe” skills and controls should be mandatory in any consumer-facing tool by 2026? 
  • Should the FSSC publish milestone deliverables (every six months) so the sector isn’t waiting until 2027 to act? 
  • Should the Treasury’s blueprint hard-wire human oversight (escalation, audit trails, liability split) into AI deployments in advice, lending and fraud, and who carries the can when AI gets it wrong? 
  • How do we avoid a London-only skills boom? What regional training models (FE colleges, bootcamps, apprenticeships) actually get smaller firms production-ready? 
  • What procurement rule changes would help banks and building societies adopt trustworthy AI faster without buying black-box risk?
     

1 responses from the Newspage community

Copy all

Star Quote
Copy

Colette Mason
AI Ethics Consultant at Clever Clogs AI
On one hand, this is exactly the kind of joined up thinking we’ve needed, getting Treasury, the City and industry around one table to define what responsible AI adoption looks like in finance.

On the other, 2027 feels like an eternity in AI years. By the time that report lands, we’ll have had several model upgrades, a few scandals, and millions more people relying on unregulated AI ‘money’ tools and homemade techniques.

This report should fast track the practical steps as soon as they’re identified, so companies and individuals can adapt in real time, not at the bureaucratic crawl of the public sector.