Do We Have to Use the 13 FSSC Defined Skills?

Most firms responding to the Skills Compact meet the same document early on: the Financial Services Skills Commission's (FSSC) 13 defined AI skills.

It reads like a checklist, so it is tempting to treat it like one — thirteen boxes to tick, thirteen training modules to build, thirteen lines of evidence to file away.

That approach is understandable, but it misreads what the framework is for.

The 13 skills exist to give the sector a shared vocabulary for AI capability, not to prescribe an identical training programme for every firm regardless of size, risk profile or workforce mix.

Understanding the difference matters, because getting it wrong in either direction — ignoring the framework, or over-engineering compliance with it — creates real operational and governance risk.

Why the 13 Skills Exist

The FSSC developed its future skills framework to solve a specific problem: financial services firms were approaching AI capability inconsistently, with no shared reference point for what "AI literate" actually means across different roles.

Without a common framework, one firm's definition of AI competence might bear no relation to another's, making it difficult for regulators, boards and the sector as a whole to assess whether the workforce is genuinely prepared for AI-enabled ways of working.

The 13 skills sit within the broader AI Skills Compact — the government-backed expectation that financial services firms actively build workforce AI capability rather than leave it to chance.

The framework's purpose is to describe the destination — the capabilities a workforce needs — not to dictate the exact route every firm must take to get there.

How Firms Traditionally Approach Sector Frameworks

This is not a new problem in financial services. The sector has long worked with non-statutory frameworks that describe expected outcomes without prescribing exact delivery — competency frameworks for conduct risk, training frameworks for regulatory knowledge, and sector-wide standards for professional qualifications.

In each case, firms have typically responded the same way: they map the framework against their own roles, identify where existing training already covers the intent, and document the gaps that need addressing.

Verbatim adoption has rarely been expected, or even practical. A retail-facing customer service team and a wholesale trading desk have very different risk exposures and operational realities, even where both need some grounding in the same underlying competency.

The FSSC's 13 skills framework should be read in this tradition — a reference standard that firms are expected to engage with substantively, evidenced through mapping and rationale rather than word-for-word replication.

Where Flexibility Genuinely Exists

Firms have real latitude in several areas.

Terminology can be adapted. A firm can rename or relabel a skill area to match its own internal language, provided the underlying capability is still addressed.

Grouping can be adapted. Several of the 13 skills often make more sense delivered together for a given role than as thirteen separate modules — a trading desk may need three of the skills at depth and only a light-touch awareness of the rest.

Depth by role can vary. A compliance officer and a junior operations analyst may both need to understand the same skill area, but at very different levels of technical depth.

Delivery method can vary. Classroom training, e-learning, structured on-the-job exposure and AI-assisted coaching tools can all be legitimate ways to build the same capability.

AI-enabled learning tools are increasingly useful here, particularly for firms adapting the framework across many role types. They can help tailor content depth and pacing to different audiences without requiring a separate manually-built curriculum for every role — though the underlying design and sign-off of what is taught should remain with the firm's own L&D and risk functions, not be delegated to a tool.

What should not change is the underlying intent and coverage of each skill area. If a firm's programme leaves a genuine gap against one of the 13 skills, relabelling or reorganising does not resolve that gap — it only obscures it.

What to Have in Place Regardless of Approach

Whatever structure a firm chooses, three things should exist and be readily available.

A documented mapping showing how the firm's own training structure relates back to the FSSC's 13 skills, even where names, groupings or delivery formats differ.

A clear rationale for any deviation — why skills were combined, why depth varies by role, why a particular delivery method was chosen — that a risk committee or auditor could review and understand.

Evidence of assessment, showing that skills have actually been built and checked, not merely that training was delivered.

Together, these three items let a firm demonstrate genuine engagement with the framework's intent, regardless of how closely its own programme structure mirrors the FSSC's original wording.

Example

A mid-sized London-based commodities trading firm is preparing its Skills Compact response.

Its L&D lead initially plans to run 13 separate training modules mirroring the FSSC skills verbatim, but finds this doesn't map well to the firm's actual roles — traders, settlement staff and compliance officers need very different depth on the same underlying skill areas.

The firm regroups the 13 skills into four role-based learning tracks — trading floor, settlement operations, compliance/risk, and senior management — and documents how each track maps back to the FSSC's original 13 skills.

This mapping is presented to the risk committee alongside assessment evidence, satisfying the intent of the framework without forcing an artificial 13-module structure onto staff who don't need it.

FAQs

• Is the FSSC skills framework legally binding?

  Not in itself — it is not statutory law. However, it sits within the broader regulatory and government expectation set by the AI Skills Compact, so it should be treated as a strong sector expectation rather than optional guidance. Firms unable to show any engagement with it are likely to face difficult questions from boards or regulators.

• Can we rename or regroup the 13 skills for our own programme?

  Yes. This is expected practice, provided the firm can show how its own structure maps back to the original framework's intent. What matters is that the underlying capability is still built and evidenced, not that the labels or groupings match exactly.

• What happens if we don't reference the framework at all?

  The firm loses a shared reference point that regulators, boards and auditors will expect to see when reviewing an AI skills response. Producing a mapping document against the framework is a relatively low-cost exercise compared with the reputational and governance risk of having no clear answer when asked how the firm's approach relates to it.