AI is already embedded across most organisations, but what is less clear is how consistently it is being used to create value. That was the central theme of the latest Corndel, Imperial Executive and Professional Education and Microsoft partnership event, where senior leaders from across L&D, HR, Data, AI & Transformation functions came together to explore a challenge many are now facing.
Despite significant investment in tools like Copilot, impact remains uneven. Across the event’s sessions, one message came through consistently from our two expert speakers: AI adoption is not being limited by technology. It is being limited by how people use it.
Stuart Munton: AI does not create value. People do
Stuart Munton, Chief for Group Operations and Technology at AND Digital, opened the session with a direct and necessary reset: “AI does not drive the value. The value comes from how people use it.”
While many organisations are still framing AI as a technology rollout, Stuart reframed it as a people and operating model challenge. He highlighted a critical point that is often overlooked in boardroom conversations:
• AI is not free
• Copilot and broader AI investments carry real cost
• That cost must be offset through measurable improvements in productivity and impact
This aligns with a broader pattern seen across UK organisations. While 97% of HR leaders report offering AI training, only 39% of employees say they have actually received it, and just 14% rate it as highly effective. Without effective capability building, the return on AI investment will always fall short. Stuart also explored the human implications of AI-enabled work. As more tasks are mediated through tools, there is a risk that collaboration becomes fragmented. He emphasised that:
• Engagement remains critical
• Teamwork becomes more important, not less
• “Human in the loop” cannot become “human working alone”
Another key takeaway centred on the nature of AI output. Left unchecked, generative AI tends towards safe, predictable responses: “AI output will be generic. Humans need to add the variability.” This is where human judgement becomes essential. The value is created not in generating content, but in shaping, challenging and applying it. Finally, Stuart addressed one of the most overlooked consequences of AI adoption. As AI enables individuals to take on broader tasks, roles begin to blur. This creates pressure on operating models:
• More people are doing a wider range of work
• Traditional role boundaries are shifting
• Organisations must rethink how work is structured and managed
Alongside this, leaders must navigate a growing set of risks. Not everything that can be automated should be. He reflects, “We need to ask not just can we, but should we.”
Emma Ledger: Why AI adoption stalls, and how to fix it
Emma Ledger, AI Workforce Go-To-Market Lead at Microsoft UK&I, built on this by focusing on why AI rollouts so often fail to deliver sustained impact. Using the ADKAR model, she broke down the common failure points in AI adoption:
• Awareness: people do not understand why AI matters
• Desire: fear of replacement reduces engagement
• Knowledge: training is too generic to be useful
• Ability: fear of failure stops experimentation
• Reinforcement: organisations fail to sustain momentum
Each of these challenges is visible in current workforce data. One-third of UK employees report feeling unprepared to adopt AI in the next 1–3 years, despite widespread access to tools. Emma then shared how Microsoft approached this internally, offering a practical blueprint for leaders:
• Awareness was driven through visible executive sponsorship and clear communication
• Desire was built through AI champions and real use-case storytelling
• Knowledge came from role-specific training and hackathons
• Ability was supported by giving people time and space to experiment
• Reinforcement was sustained through measurement, using tools like Viva Insights and Power BI to track usage and impact
The emphasis reflected that adoption is not a single intervention. It is a structured change process that requires ongoing reinforcement.
Event Key Takeaways: From AI investment to AI impact
Across both sessions, a clear picture emerged of how organisations are progressing with AI and where they are getting stuck.
The biggest barrier to realising value is no longer access to technology, but capability. While many organisations are moving from experimentation to deployment, progress remains uneven because workforce readiness has not kept pace with investment.
Even where AI is already driving productivity gains, value is not guaranteed. Creating capacity is only the first step. Leaders must make deliberate decisions about how that time is reinvested. This requires a sharper focus on redesigning workflows, managing time at a team level, and building both technical and cognitive skills. As a result, capabilities such as critical thinking, creativity, ethics and empathy are becoming more important, not less.
At the same time, generic training continues to limit impact. There is a clear disconnect between the volume of training delivered and its effectiveness. Programmes that are not role-specific or grounded in real work are unlikely to drive behaviour change. As highlighted through the ADKAR model, capability must be built in context to translate into adoption.
AI is also reshaping operating models, not just individual tasks. As people use AI to expand the scope of their work, organisations are seeing blurred role boundaries, shifts in decision-making, and an increased need for governance and ethical oversight.
Ultimately, culture is the factor that determines whether adoption is sustained. While technology can be deployed quickly, behaviour change takes time. Organisations seeing the most progress are those that encourage experimentation, set clear guardrails, and continuously measure and iterate their approach.
Organisations that succeed will treat AI as a capability to be built and embedded across the business. AI alone does not deliver value; it requires leadership clarity, workforce capability, and cultural alignment.
As Stuart summarised: “Transformation is not about technology. It’s about how people use technology to create value.”