To read all the insights and benefit from all learnings, join DataIQ and access our members only Hub.
Move training courses into experiential learning
Several leaders argued that relying on formal learning platforms cannot build AI capability quickly enough. For success, these platforms need to be combined with structured learning with practical programmes where practitioners solve genuine business problems alongside experienced mentors.
One organisation had run internal cohorts where small teams work on real business challenges using enterprise data. Successful leaders then become coaches for the next cohort, creating a self-sustaining learning model that continuously grows internal capability. This approach exposes practitioners to data governance, ownership, and responsible AI in practice.
Communities drive adoption faster than central teams
Organisations have established networks of AI champions or ambassadors embedded across business units. Instead of relying on small central AI teams, these local advocates encourage adoption, share reusable prompts and use cases, celebrate successful projects, and surface practical challenges from frontline teams.
Leaders found this significantly more effective than top-down communication, alongside monthly showcases and internal recognition programmes, because learning spreads through trusted peers.
Technical debt becomes a people problem
GenAI has drastically reduced the time needed to build prototypes, but leaders warned that rapid experimentation is creating a new challenge. Without governance, organisations risk accumulating hundreds of local tools that cannot scale which increases technical debt while consuming scarce engineering resources.
Rather than restricting innovation, leaders want to introduce stage gates that move ideas from prototype to proof of concept, proof of value, and finally production. This shifts the emphasis to building AI that delivers sustainable business value.
External learning has limits
While platforms were widely used for foundational learning, peers agreed they struggle to keep pace with emerging AI technologies.
To address this, organisations increasingly supplement external content with internally developed learning materials, expert-led sessions, short webinars, and curated learning pathways tailored to their own technology estate and business context. As one leader observed, external platforms are “very good, but they’re quite generalist, and we wanted to ground it with real examples of our business.” This leader then built the solution internally.
Career progression cannot rely on people management
Leaders questioned traditional technical career paths, where advancement usually means managing larger teams. Instead, organisations are beginning to create broader pathways by giving technical specialists opportunities to lead innovation, rotate across functions, work internationally, or build enterprise-wide AI capabilities without becoming line managers.
Leaders acknowledged this remains an unresolved challenge but agreed that retaining specialist talent will depend on offering progression without forcing management responsibility.
Demand for AI engineering exceeds supply
Despite discussion of future skills, leaders consistently identified immediate shortages in AI engineering and data engineering. As organisational appetite for AI accelerates, organisations have shifted from encouraging adoption to struggling to resource it. Competition for experienced practitioners remains intense, particularly outside large technology firms.
Leaders are responding through internal development rather than relying on recruitment, while some are expanding graduate and apprenticeship programmes to grow future capability from within.
To read all the insights and benefit from all learnings, join DataIQ and access our members only Hub.
