From Experiment to Enterprise: What CTOs Are Learning About AI and Social Mobility
What will we know about AI tomorrow that we wish we had acted on today?
That question sat at the centre of a recent Progress Together CTO Roundtable, held in association with Grant Thornton UK, where senior technology leaders from across financial services came together to explore how artificial intelligence is reshaping organisations and where it risks reinforcing existing inequalities if handled carelessly.
The conversation moved deliberately beyond tools and pilots, focusing instead on governance, workforce impact, and the role of technology leaders in shaping fairer outcomes.
Still in the experiment phase – but patterns are emerging
Participants were candid that, despite widespread access to generative AI tools, most organisations remain firmly in an experimental phase. While many colleagues now have access to Copilot or chatbot tools, the leap from individual productivity to enterprise-wide, customer-facing deployment has been slow.
This is partly a question of provider maturity. But just as important are the internal governance mechanisms around the technology. Highly regulated firms already have strong governance muscle memory – controls, audit trails and risk escalation are embedded in their culture. Many boards are intentionally cautious, particularly in sectors where explainability and auditability are non-negotiable.
A recurring theme was the question of human oversight and control. Several organisations described deliberately restricting what AI systems can do, and ensuring humans remain both driving and accountable for outcomes.
A people and business change
A consistent refrain across the discussion was that AI adoption fails when treated as a purely technical rollout. The organisations seeing the most value are those investing heavily in training, engagement and cultural change.
Most firms represented have now trained all colleagues in responsible AI use, with targeted sessions for senior leaders. Several are actively tracking who has access to which tools, how embedded AI use is in day-to-day work, and how training affects confidence and adoption.
Importantly, colleagues who are slower to adopt are not automatically viewed as resistant. In some cases, those asking the toughest questions are those with the strongest professional standards or the greatest awareness of potential customer impact.
Reverse mentoring has also emerged as a powerful enabler – pairing senior leaders with colleagues who are more fluent in AI tools to accelerate learning across the organisation.
Early evidence points to inclusion benefits
Some of the most striking insights from the roundtable came from examining who benefits most from AI tools in practice.
Participants shared early evidence that part-time colleagues – historically slower to progress – are saving up to 70 minutes a day using AI, reducing the catch-up burden and freeing time for learning. Neurodivergent colleagues are reporting significant improvements in wellbeing, confidence and performance, particularly through writing and communication tools. And colleagues who previously self-selected out of progression opportunities are now more likely to put themselves forward.
Early engagement data from some organisations also suggests particularly high uptake among colleagues from ethnic minority backgrounds, though participants were cautious about drawing firm conclusions without more robust workforce data.
The risk of leaving people behind
While enthusiasm for AI was clear, so too were the anxieties about who might lose out.
The discussion highlighted that in some organisations, people are spending their own time experimenting with new technology – which is easier for some than others. More broadly, participants warned that unequal access between teams, roles or organisations could just as easily widen pay and progression gaps if left unchecked.
AI has the potential to make progression more transparent and controllable. But only if organisations are intentional about how tools are rolled out and supported.
Recruitment practices are already being adapted in response. Some organisations are running assessment exercises twice – once with AI and once without – or moving back towards interviews and situational judgement, recognising that AI-assisted applications are easier to game in artificial settings.
Sustainability and geopolitics
Later in the discussion, attention shifted to the wider consequences of AI adoption. Participants acknowledged growing concern among colleagues and customers about the environmental footprint of AI, particularly energy and water usage. Some employees are actively opting out of AI tools for sustainability reasons, and organisations stressed the importance of transparency about environmental impacts, even where answers remain incomplete.
Geopolitics also featured prominently. Reliance on overseas providers operating under differing regulatory regimes, and the ethical maturity of relatively young AI vendors, raise strategic questions for technology leaders selecting long-term partners.
From defence to opportunity
Most organisations represented described themselves as currently playing defence – focused on doing no harm to customers or colleagues while the technology matures. Yet there was a shared recognition that this will not be enough.
“The real challenge ahead is shifting confidently from risk mitigation to responsible opportunity – using AI to remove structural barriers, support diverse career paths, and prepare the workforce for profound change. Some organisations are already at this stage, and it is essential that learning is shared across the sector.”
Sophie Hulm, CEO of Progress Together
What leaders are focusing on now
The reflections from the roundtable point to a number of emerging priorities for firms navigating AI adoption:
- Scaling AI with strong governance, oversight and accountability – moving from experimentation to prioritised, high-value use cases with clear maturity frameworks
- Investing in people – building organisation-wide digital capability and driving the cultural change needed for responsible adoption
- Ensuring equitable access – proactively managing access to AI tools and providing targeted support to prevent widening gaps in productivity, progression and opportunity
- Using AI to advance inclusion – scaling approaches that are improving outcomes for underrepresented groups and embedding AI capability into progression pathways
