🏠︎ | Past Sessions | From Insight to Impact AI & Data Upskilling in Action
- Event: Finance Forum 25
- Date: 7 October 2025
- Speakers:
- Josh Hubbert, CEO, Workforce Learning, QA
- John Pochribniak, Learning and Development Manager, AXA UK
- Estimated read time: 7-8 minutes
Quick summary
- AI impact is universal, but skills, not tools, determine whether value is realised
- Role specific learning beats generic training, especially for finance and leadership roles
- Human skills matter more, not less, particularly judgement, critical thinking and decision making
- Adoption depends on trust and governance, not just access to technology
- Communities and champions turn learning into behaviour, not just capability
Skills, not tools, define AI impact
The session opened with a clear challenge. As AI accelerates, the differentiator is not access to technology, but whether organisations invest seriously in skills. Josh Hubbert argued that every role is now affected by AI, including finance, and that this makes upskilling more important, not less.
He pushed back on the idea that AI replaces human value. Instead, the more technology advances, the more organisations rely on distinctly human capabilities such as judgement, context, and critical evaluation. This is especially relevant in finance, where outputs must be interpreted, challenged, and translated into decisions.
Alongside technical capability in data and AI, Hubbert highlighted growing demand for leadership skills and critical thinking. Prompting alone is not enough. Teams must be able to assess outputs, recognise limitations, and iterate responsibly.
“In the age of AI, skills become more important than ever.” Josh Hubbert, CEO, Workforce Learning, QA
Role specific learning is replacing one size fits all training
A major theme was the shift away from generic AI awareness programmes towards role specific pathways. Hubbert described a clear trend, organisations now ask how AI applies to a specific function, such as finance, rather than treating AI as a standalone topic.
QA has responded by designing targeted learning pathways aligned to real roles. This includes finance professionals, technical specialists, leaders, and emerging AI champions. The intent is practical relevance, not abstract capability building.
Another shift discussed was in learning formats. While digital self paced learning remains valuable, there has been renewed demand for expert led and blended learning. Participants want interaction, discussion, and the ability to ask questions in context, particularly for complex or high risk topics such as AI use in regulated environments.
Using the levy to fund data and AI upskilling
Hubbert also drew attention to a practical but often overlooked lever, the apprenticeship levy. Many organisations pay into the levy but do not fully use it, meaning funding returns unused to government.
He emphasised that the levy can support data and AI upskilling, including for mid career professionals, not only early career entrants. In practice, this allows organisations to reskill existing talent and build capability at pace without relying solely on short courses.
The session reinforced that apprenticeships should be viewed as a strategic investment. They require time and commitment, but deliver deeper skill uplift and longer term value.
AXA’s Data and AI Academy, structure before scale
John Pochribniak shared how AXA UK translated these principles into action through its Data and AI Academy. Launched in 2024, the academy brought together previously fragmented learning initiatives into a single, coherent structure.
The academy is built around four pillars, capability, literacy, collaboration, and talent. Each pillar addresses a different challenge, from developing deep technical skills in data professionals, to building baseline data literacy across the wider workforce.
A key early step was definition and discovery. AXA invested time understanding the day to day reality of different roles, then aligned learning pathways to strategic objectives. This ensured that development activity was grounded in real work, not abstract frameworks.
Literacy, collaboration and talent as force multipliers
Beyond capability building, AXA focused heavily on literacy and collaboration. For roughly 9,000 employees, literacy initiatives aim to create shared understanding of data and AI tools, how they work, and what responsible use looks like.
Collaboration addressed a common learning challenge. Development activity often sits outside the business. AXA deliberately embedded learning into live business problems, working with leaders to ensure relevance and momentum.
Talent was the final pillar. With demand for data skills increasing, AXA focused on retention, progression, and employer attractiveness. The academy became part of the employee value proposition, signalling that learning and growth are ongoing, not episodic.
Driving adoption of generative AI in practice
A central example was AXA’s internal generative AI tool, Secure GPT. Access alone did not guarantee impact. Adoption required confidence, clarity, and governance.
AXA supported this through multiple initiatives, including practical clinics, prompt sharing, and guidance on what information could and could not be used. Monthly expert led sessions allowed staff to ask real questions tied to their work.
Use cases were deliberately grounded in operational reality. One example was call summarisation for customer facing teams, helping reduce manual effort while improving accuracy. The same tools were then made available to other functions, including finance, to explore productivity gains in context.
“If people do not trust how to use it, they will not adopt it.” John Pochribniak, Learning and Development Manager, AXA UK
Communities and champions turn learning into behaviour
One of the strongest messages was the role of communities. AXA established communities of practice across areas such as Secure GPT, data professionals, and Power BI, hosted on internal collaboration platforms.
Initially driven centrally, these communities evolved into two way spaces where employees ask questions and experts respond in near real time. This peer interaction proved critical in embedding learning into daily work.
Alongside communities, AXA developed a Gen AI champion model, supported in part by a level three apprenticeship. Champions receive deeper training, then act as local advocates, coaches, and feedback loops across the organisation.
This approach created visible ownership of AI adoption, reducing reliance on central teams and accelerating practical experimentation.
What this means for finance leaders
For finance leaders, the session reinforced that AI and data upskilling is not a technology project. It is an operating model change that affects capability, behaviour, and decision making.
Questions to ask your organisation- Are our AI and data programmes role specific, or generic
- Do finance teams have space to practise and ask questions safely
- Are leaders equipped to judge outputs, not just sponsor tools
- Are we using levy funding strategically for data and AI skills
- Whether adoption is spreading beyond early adopters
- Whether communities are active without constant central push
- Whether AI use cases are tied to real finance workflows
- Whether governance enables confidence rather than fear
- Treating AI training as a one off awareness exercise
- Assuming access equals impact
- Ignoring human and leadership skills alongside technical learning
- Centralising ownership without building local champions
What good looks like
Effective AI and data upskilling combines structure with flexibility. Finance teams understand how tools apply to their roles, leaders can judge and challenge outputs, and communities support continuous learning. Technology becomes an enabler of better decisions, not an end in itself.
Conclusion, from insight to impact requires deliberate design
The session made a clear case that AI value is unlocked through people, not platforms. Organisations that invest in structured, role relevant learning, supported by communities and champions, move faster from experimentation to impact.
For finance leaders, the opportunity is to shape how AI and data skills are built inside the function, ensuring that productivity gains translate into better insight, stronger judgement, and more confident decision making.
Speakers
CEO, Workforce Learning, QA. Josh leads QA’s workforce learning business, with prior experience as COO at GfK and senior roles in strategy, M&A and management consulting.
Learning and Development Manager, AXA UK. John leads AXA UK’s Data and AI Academy, focusing on data literacy, capability building and the development of data professionals.