When a sector's own skills authority publishes research that runs counter to most of the sector's training assumptions, it is worth reading it carefully. That is what the Financial Services Skills Commission did in May 2025 with Unlocking AI's Potential: The Skills That Matter.
This article sets out what the report actually says - the specific data points, the specific role-by-role findings, the specific recommendations - and what it means in practice for financial services L&D strategies.
The FSSC cites Department for Education analysis assigning financial services an AI occupational exposure score of 1.1, against an all-industry average of 0.2. Financial services has the highest AI exposure score of any sector in the UK economy. The sector is also the most exposed to large language models specifically - the GenAI tools that have driven the most rapid adoption since 2022.
What this means for L&D: AI upskilling is not optional in financial services and it cannot be deferred. The sector faces more AI-driven role change than any other - which means the capability development required to navigate that change is more urgent, more widespread and more consequential than in other sectors. The question is not whether to invest in AI capability development but how to do it effectively across the full workforce.
The FSSC's most important segmentation finding is this: only 0.5% to 1.5% of the financial services workforce needs specialist AI skills. Those who implement AI technology constitute approximately 10% to 15%. The vast majority - 85% to 90% - need AI understanding (what AI is, how to use specific tools responsibly and effectively) and the supporting human skills and behaviours that make AI adoption successful.
What this means for L&D: most AI training budgets are disproportionately invested in the 1-2% of the workforce that needs specialist skills - through recruitment strategies, technical certifications and data science programmes. This is necessary but insufficient. The 85-90% who need AI understanding and human capability are typically underserved - and they are the population that determines whether AI adoption becomes organisational performance or remains departmental experimentation.
The FSSC's member survey and vacancy data analysis from Simply Get Results (including Lightcast data) produce a finding that should be central to any FS L&D strategy. The supply-demand gap for creative thinking and adaptability is 35 percentage points - the same as the gap for AI and machine learning. The gap for coaching is 30 percentage points. Relationship management and empathy appear in more FS job postings than most specialist AI technical skills.
The FSSC frames this explicitly: "Although workers across our sector will need to increase their knowledge of technology and its application, it is their behaviours - such as empathy and relationship management - which will unlock the full potential of AI."
What this means for L&D: programmes that develop technical AI knowledge without simultaneously developing the human skills the FSSC identifies as equally undersupplied are not building the integrated capability the sector needs. They are addressing the surface of the problem, not its foundation.
The FSSC's role-by-role analysis, drawing on EY Skills Foundry vacancy data from 2019 to 2024, shows how the skills demanded in individual financial services roles are shifting in response to AI:
Financial analysts and advisers: decision-making up nine places year on year in the skills demand ranking. Process improvement and optimisation up significantly. Microsoft Excel declining. Writing and editing declining. The analyst role is shifting from technical production to analytical judgement. Development programmes for financial analysts should reflect this - investing more in decision-making, analytical reasoning and stakeholder communication, and less in the technical skills that AI is increasingly handling.
Project managers and account managers: listening and interpersonal skills entering the top 25 demanded skills. Microsoft Excel declining by four to nine places. Writing and editing declining by seven to nine places. Communication techniques and practice rising. The project and account management role is becoming more relational, less administrative. Development programmes should build facilitation, communication and relationship management capability alongside any technical AI content.
Financial services managers and directors: process improvement and decision-making consistently rising. Leaders in control functions face new challenges around AI governance, hallucination management and oversight accountability. Development programmes for leaders need to include AI governance and responsible use alongside the communication and stakeholder capabilities needed to lead AI-enabled teams.
HR managers and HR administrative roles: the FSSC identifies these as among the roles most impacted by large language models. For HR managers, influencing has risen to fourth in the skills demand ranking, accountability has risen 32 places, coaching has risen five. Technical reporting skills are declining. The HR function needs the human skills to lead AI-enabled change at the same time as it is managing the technical demands of AI adoption. This is the specific context for which the FTLP programme was designed.
The FSSC's data on training provision reveals a significant mismatch between the scale of the AI skills challenge and the current response. Only one in five financial services professionals receives AI training (ONS data). Only 25% of companies plan to offer GenAI training (Microsoft/LinkedIn survey). Only 18% of firms integrate AI training into graduate programmes (EY survey).
81% of firms experience lack of specialist talent as a barrier to AI adoption - but the FSSC is clear that the specialist talent problem affects only a small fraction of the workforce. The broader workforce capability problem - the gap in AI understanding and human skills that affects 85-90% of colleagues - is the larger issue, and it is less addressed.
What this means for L&D: the framing of the AI skills challenge needs to change. It is not primarily a recruitment challenge for rare technical talent. It is a development challenge for the broad workforce - one that requires systematic investment in AI understanding, human skills and management capability at scale, using the levy funding mechanism that already exists.
The FSSC's conclusion and recommendations section (page 30 of the report) sets out three actions for employers. These are worth stating directly.
Prepare for AI adoption by assessing your skills needs. Segment the workforce, assess what level of AI skills are needed for each group, and identify where the gaps are. The FSSC's Skills Gap Analysis Toolkit can help firms make faster progress.
Build targeted training programmes. The ONS data shows only one in five individuals receives AI training - firms should focus on building relevant training to enable effective AI implementation. The FSSC's updated Future Skills Framework provides the tool to identify the training required for different groups of colleagues.
Collaborate to explore how AI can support a skills-based approach. Case studies from FSSC members show how AI-powered tools can drive skills forecasting, learning and internal mobility - and thereby close the skills gap over time.
An FS L&D strategy that genuinely reflects the FSSC's findings looks different from most current practice in three ways:
It addresses the full 85-90%, not just the 1-2%. It invests systematically in AI understanding and human capability development for the broad workforce, not only in specialist AI skills recruitment and development for a small technical cohort. It uses the levy - which is already paid - to fund development that would otherwise require direct budget: the Level 3 and Level 4 AI programmes, the data programmes, the management and transformation programmes that build the integrated capability the FSSC identifies as most needed.
It treats technical and human capability as inseparable. Not sequential, not separate budgets. Integrated development - AI understanding alongside communication, stakeholder engagement, judgement and adaptability - in programmes applied to real FS work.
And it is role-specific. The development that financial analysts need is different from what project managers need, which is different from what HR managers need. The FSSC's role-by-role analysis provides the evidence base for that differentiation. L&D strategies that reflect it will be more effective, more efficient and more aligned to the real skill shifts the sector is experiencing.
FTL's financial services programmes are built directly on this framework - developing integrated technical and human capability, for specific FS roles and work contexts, fully funded through the apprenticeship levy.
To explore the right programme mix for your organisation: talk to the FTL team
For a deeper look at the FSSC evidence and what it means for mid-sized FS firms, read our whitepaper: The human skills imperative in financial services: what the FSSC research demands from L&D leaders