How can organisations measure the success of their AI initiatives and what remains exclusively the domain of the human worker?
Rakshit: I tend to avoid measuring AI success by how many tools have been deployed, or even just by hours saved. Saved from what? Did that work actually disappear, or did it move into checking outputs and managing exceptions?
Organisations should also be measuring whether AI is helping teams make decisions faster, reducing operational complexity, improving employee experience, lowering support demand and delivering meaningful business outcomes. I’ d look at whether employees are using AI with confidence, whether support demand is falling, whether onboarding is smoother, and whether risks are easier to identify and address. The outcome should be better work, not just more technology in the stack.
Trust, governance and accountability must be included in that equation too. Lenovo research shows that 90 % of IT leaders see gaps in their ability to defend against AI-driven threats, so success can’ t just mean faster adoption. It has to mean adoption the business can stand behind. What stays human is judgement and accountability. AI can recommend and automate, but people still decide what good work looks like.
64 October 2026