Matthew Skelton's argument at QCon London 2026 is the one that too many organisations are avoiding: AI success is not a technology problem, it is a maturity problem. He is right. The companies that see real returns from AI are not the ones with the most advanced models or the biggest budgets. They are the ones that have figured out how to manage agency, security, and stewardship at scale. For anyone building or buying AI tools, this reframes the conversation entirely. You cannot bolt AI onto a chaotic organisation and expect transformation. You have to prepare the ground first.
What Skelton calls "bounded agency" is the practical heart of the matter. Giving AI agents decision-making power without clear boundaries is a recipe for drift, not productivity. The same principle applies to human teams. If your internal processes are siloed, your AI tools will inherit those silos. Skelton's focus on Innovation and Practices Enabling Teams is a direct challenge to the common approach of dropping a new tool into a team and expecting it to spread organically. Knowledge diffusion does not happen on its own. It requires deliberate structure, dedicated teams, and a stewardship model that treats AI agents as something to be guided, not just deployed.
This is not abstract theory. It translates directly into how you build your data workflows. If your spreadsheet experience feels constrained by legacy tools, the path forward is not a more powerful spreadsheet. It is a more mature organisational approach to data management. AI can simplify complex tasks, but only if the underlying systems are designed to accept that simplification. Bounded agency means your AI agents know what they are allowed to do. Stewardship means someone is responsible for their output. Security means the data they touch is governed by policy, not guesswork. These are not technical hurdles. They are organisational ones.
The practical takeaway is this: before you invest in another AI tool, invest in the maturity of your teams. Map your processes. Define where agency begins and ends. Build enabling teams whose job is to spread knowledge, not hoard it. The real return on AI comes when your organisation is ready to absorb it. Skelton's framework gives you a way to measure that readiness. Use it.