3 best practices for CIOs to assess AI ROI

CIOs are under pressure to prove AI value, but many organizations still lack a clear process for measuring ROI, especially as AI spend starts to behave like a FinOps problem rather than a pure innovation play. The article argues that IT leaders should manage AI as a portfolio, balancing quick wins with strategic bets, and evaluate initiatives with full cost-benefit models that account for not just model/token costs but also business outcomes, trust, and operational efficiency. For IT organizations, this means shifting from tracking activity to reporting measurable results across the AI lifecycle so they can scale funding, govern risk, and prioritize investments that deliver both productivity and revenue impact.