What happened
McKinsey (QuantumBlack / Tech & AI) published this quantitative analysis on 2026-10-14 introducing its proprietary 'AI value curve' tool, which compares 'the cost of a task as performed today with the full cost of doing it with AI, revealing a roughly $2.5 trillion annual global opportunity in deploying the technology across certain sectors, regions, and occupations.' It frames the stakes via $7 trillion of AI infrastructure investment by 2030 needing roughly $760 billion in annual profit pools to be justified. Three patterns stand out: the strongest economics come from augmenting high-skilled, higher-wage work rather than large volumes of lower-wage work; substantial value exists in sectors beyond tech and financial services, including healthcare and education; and the same use case can pay off in one market but not another. The authors argue most organizations 'are still measuring AI adoption instead of value' and that leaders must focus investment where the economics are strongest, manage costs alongside adoption, and plan to redeploy the people AI frees up.
Why it matters
Provides a defensible unit-of-analysis for justifying AI ROI to the board: a per-task full-cost comparison that shows augmenting high-skill work, not automating cheap labour, is where the value is — directly informing where executives should (and should not) fund AI initiatives.
Action needed
Use the AI value curve framing to build a task-level ROI screen for the FY27 AI portfolio and challenge initiatives whose business case only counts adoption/token metrics rather than full cost of task delivery.