Everyone is talking about AI strategy. Far fewer teams are talking honestly about AI skills. A 2025 report found that 65% of organisations have had to abandon AI projects because they did not have the skills to support them. That is not an AI problem. It is a people and capability problem. If you work in infrastructure, cloud, DevOps, security or IT management, you are probably seeing this already: AI tools are being bought faster than people are being trained Projects are scoped without involving the people who will run them MSPs and legal firms are asked to "add AI" to services without giving their teams time to upskill Our take: AI adoption should look less like a shopping list of tools and more like a skills roadmap. The organisations that are winning are doing three things well: Defining the skills they actually need across architecture, data, security and operations Investing in upskilling and hiring in parallel, rather than expecting existing teams to absorb everything Being realistic about where partners add value, whether that is an MSP, a specialist vendor or a recruitment partner If your AI plans for 2026 are ambitious, a good first question is not "which model should we use" but "do we actually have the people who can make this live inside our environment and keep it safe".
Why AI adoption fails: a people and capability problem
This title was summarized by AI from the post below.
That 65% figure cuts across sectors, from MSPs and SaaS providers through to regulated industries like finance and legal. The common pattern is the same: tools arrive faster than skills. If you are mapping your AI roadmap and unsure which skills to build in house and which to bring in from outside, we are happy to share what infra, cloud, security and data leaders in our network are doing in practice.