In all sectors, organisations have spent a lot on artificial intelligence, but a significant percentage of these projects fail to yield tangible results. This article proposes that this is not due to the technology per se, which has developed quickly, but to the skill deficits of the leaders of such technology adoption. All too frequently, senior technology leaders view AI as a product that can be bought and turned on, rather than a transformation programme for the entire enterprise that requires disciplined governance, data readiness, workforce engagement, and financial accountability. Based on the experience of 30 years of seeing what works and what doesn't work in manufacturing, global capability centres and large enterprises alike, the discussion highlights common problems pilots without success criteria, disjointed data, legacy systems that do not integrate, front-line workers who are not engaged with new tools, and a lack of ability to prove return on investment. Next, it outlines the disciplines of successful adoption defined problem, executive sponsorship, governed data, realistic integration, human oversight, and continuous measurement. The article also looks at the development of specific leadership positions. The key message is that positive businesses thrive when AI is viewed as a business change initiative led by smart, self-directed business leaders, rather than as a procurement decision handed over to vendors. This article is aimed at executives, practitioners and the interested public.
Publication Date: 2026-06-25