
What It Really Takes to Build AI Capability Across an Organisation

Key Takeaways
- •88% of firms use AI in at least one function (2025).
- •Only ~33% have moved beyond pilot projects to scale AI.
- •Training alone doesn’t create AI capability; workflow redesign is essential.
- •Six interdependent components drive organizational AI capability.
- •Leaders must treat AI as a system‑design challenge, not HR problem.
Pulse Analysis
The rapid diffusion of artificial‑intelligence tools has turned curiosity into a mainstream business function. McKinsey reports that 88% of companies now deploy AI in at least one area, and employee usage has jumped from 30% to 76% in just two years. Yet the data also reveals a stark capability gap: only a third of firms have progressed past experimental pilots, and the majority lean on large‑scale training initiatives that fail to move the needle on performance.
Why does training fall short? Individual tool fluency—knowing how to prompt ChatGPT or use Copilot—does not automatically reshape the way work gets done. Without redesigning workflows, updating quality‑control structures, clarifying decision rights, establishing continuous learning loops, revising hiring and promotion criteria, and granting cultural permission to experiment, AI‑generated output collides with legacy processes, creating bottlenecks and eroding trust. These six interdependent elements form the backbone of true organisational AI capability, turning isolated tool use into coordinated, value‑creating operations.
For leaders, the prescription is clear: stop treating AI as an HR or training problem and view it as a system‑design challenge. Start by selecting a single high‑impact workflow, map how AI alters each step, and redesign downstream hand‑offs, reviews, and decision points accordingly. Embed continuous, peer‑led learning into the work itself, and align hiring and promotion metrics with the new AI‑enhanced roles. Companies that master this holistic approach within the next 18 months will convert AI spend into tangible productivity gains and secure a durable competitive edge.
What it really takes to build AI capability across an organisation
Comments
Want to join the conversation?