AI Pilots Worked. Scaling Them Into Regulated Operations Is a Different Problem
Most UK and European financial services firms have run AI pilots. The proof of concept is done. The harder part moving AI safely into live, regulated operations is where firms are now getting stuck. And it is almost never the technology causing the delay.
60%
2023–2024
rise in demand for model risk and AI governance roles in UK financial services,
Why is moving from AI pilot to live operations so difficult?
Running a pilot is relatively straightforward. A small team, limited risk exposure, room to learn.
Moving into live regulated operations is a different challenge entirely.
Every AI decision needs a named owner. Every model needs documented human oversight. Every deviation needs a recorded trail, one that will hold up if the FCA or an EU regulator asks to see it.
The technology is not the hard part.
The governance and people structure around it is. And most workforce plans were not built for this combination of pace and regulatory obligation.
What is 'shadow AI' and why does it matter?
Shadow AI refers to AI tools that employees are already using at work, without formal sign-off from risk, compliance or IT teams. It happens across underwriting, claims, finance and operations.
It is not always malicious; often people are just trying to do their jobs faster. But it creates real compliance exposure: decisions being made by AI that nobody in governance has approved or documented.
Shadow AI is already inside your organisation
Most financial services firms have shadow AI, they just don't know the full extent of it.
People in underwriting, claims, finance and risk are using AI tools that compliance teams have limited visibility of. The tension between moving fast and staying compliant is real and it doesn't resolve itself without someone actively managing it.
The question isn't whether shadow AI exists in your firm. It almost certainly does. The question is whether you have the right person to govern it without bringing innovation to a halt.
The firms managing this well appoint someone who can design guardrails that protect the firm without blocking the work. That is a rare combination: regulatory knowledge, operational credibility and the authority to hold the line between both.
Workforce Readiness for AI in Financial Services
Get the Whitepaper and Score Your Organisation
Who should own AI governance in a financial services firm?
This is one of the most common questions we hear and one of the most poorly answered inside most firms.
AI governance has often drifted to whoever raised their hand: an IT team, a compliance function, a transformation office. That rarely works at scale.
AI governance needs to sit close to CEO or COO level. It needs enterprise-wide authority, the ability to drive change across risk, compliance, technology and the business at the same time. Without that authority, it becomes advisory at best.
In insurance specifically, the challenge is sharper. Actuarial governance frameworks are well established but they weren't built for AI trained on dynamic data at scale. The people who understand both are genuinely hard to find.
How long does it take to hire for these roles?
Three to six months is realistic for the profiles that matter most right now.
- Model risk managers with practical AI experience
- AI governance leads who understand FCA compliance and can work alongside technologists
- Change directors with experience managing DORA obligations alongside live AI programmes
Firms that are mid-programme and facing a gap today are already behind. Emergency hiring under pressure costs more and produces worse results than planned hiring does.
The firms ahead started building these pipelines before they felt urgent and that is exactly why they're ahead.
Workforce Readiness for AI in Financial Services
If this raises questions about your own position, the full whitepaper goes further.
It covers the Execution Readiness Diagnostic, scorecard, readiness profiles and seven strategic questions.
Frequently Asked Questions
What is shadow AI in financial services?
Shadow AI is AI tools that employees use at work without formal approval from risk, compliance or IT teams. It is common in underwriting, claims and finance and creates governance exposure most firms have not formally addressed.
Why is moving AI from pilot to production difficult in financial services?
Live regulated operations require named accountability, documented human oversight and audit trails that hold up to regulatory scrutiny. These are organisational and people challenges not technology ones.
Where should AI governance sit in a financial services firm?
Close to CEO or COO level, with a named senior owner who has genuine enterprise-wide authority and a team to act on it. Delegating to IT or compliance without that mandate rarely succeeds at scale.
What roles are hardest to hire for AI governance in UK financial services?
Model risk managers with commercial AI experience, AI governance leads who bridge innovation and FCA compliance, and change directors with DORA delivery experience, all with typical hiring timelines of three to six months.
--------------------
We're committed to enabling our clients and candidates to thrive in a constantly changing world. By helping businesses grow and supporting people in leading more fulfilling lives both at work and beyond, we aim to make a lasting impact. Our ultimate goal is to become the world’s most valued and trusted workforce solutions company, trusted by our people, our clients and our communities.
Our latest insights
