McKinsey’s new AI report argues the productivity payoff is real but conditional

McKinsey's new report identifies an 'AI paradox': enterprise AI adoption is growing and capital investment is accelerating, yet sustained productivity gains remain elusive. The core finding is that most companies are using AI to accelerate existing workflows rather than redesigning processes around AI — analogous to early factories that replaced steam engines with electric motors without rethinking the factory layout. McKinsey's own deployment (25,000 AI agents alongside 40,000 consultants, targeting 1:1 parity by year-end) lends credibility to its argument, but also raises questions about whether the firm's framing is calibrated to sell transformation engagements. The report lands amid skeptical data: 56% of CEOs in PwC's survey report getting 'nothing' from AI investments, 80% of companies in an NBER study report no productivity impact, and JPMorgan warns current AI infrastructure requires $650B in annual revenue to justify returns. McKinsey argues value will accrue to the ~6% of companies that redesign workflows, scale fast, and embed AI into core processes — not those that bolt AI onto existing operations.