McKinsey's strategy practice has released a new analysis arguing that the corporate world faces an 'AI paradox': adoption of generative and agentic AI is growing rapidly, capital investment is accelerating, yet sustained impact on performance remains elusive. The report, titled 'AI productivity gains and the performance paradox', contends that most current AI applications are tools that accelerate existing work but largely preserve underlying workflows. Larger productivity gains, the firm argues, will only emerge once organizations redesign processes around AI rather than simply bolting it on top.
The Historical Lesson
The report draws a powerful analogy to the introduction of electricity in factories. When electric motors first arrived, many businesses simply replaced steam engines, capturing efficiency gains but leaving the traditional line-shaft layout unchanged. The breakthrough came later, when small motors enabled managers to rearrange machines around workflows, eventually leading to entirely new operating models. McKinsey emphasizes that general-purpose technologies rarely create value in a single wave.
The Skeptical Evidence
The report lands amid growing evidence of a gap between AI investment and measurable returns. The Federal Reserve Bank of St. Louis has measured only 1.9% excess cumulative productivity growth since ChatGPT launched in November 2022, a figure well below the rates required to justify current AI capital spending. JPMorgan published a capex analysis warning that $650 billion in annual revenue would be needed 'into perpetuity' to deliver a 10% return on current AI infrastructure investment, drawing a direct parallel to the late-1990s telecom fiber buildout.
MIT Media Lab research found that 95% of organizations see no measurable returns from AI adoption. Deloitte's 2026 'State of AI in the Enterprise' report, surveying over 3,200 leaders, found that 66% report productivity gains but only 20% report revenue growth, and only 34% are using AI to deeply transform products or processes. PwC's 2026 Global CEO Survey found that 56% of CEOs say they have gotten 'nothing out of' their AI investments, and only 12% report AI both growing revenues and reducing costs. Workday's 2026 research revealed that 37–40% of the time AI supposedly saves is consumed by reviewing, correcting, and verifying AI-generated output. A February 2026 NBER study found that 80% of companies actively using AI report no productivity impact at all.
The macro picture is further complicated by a divergence in expert estimates. McKinsey itself has previously estimated AI could add $4.4 trillion to the global economy, while Nobel laureate Daron Acemoglu projects modest 0.5% productivity gains over the next decade. The hundredfold gap between these figures defines the uncertainty within which every enterprise AI capital allocation decision is being made.
McKinsey's Own Deployment
What gives McKinsey's skeptical framing particular force is the firm's own aggressive AI deployment. CEO Bob Sternfels stated at CES 2026 that the firm runs 25,000 AI agents alongside its 40,000 human consultants, and expects to reach 1:1 parity, 40,000 AI agents, by year-end. McKinsey saved 1.5 million hours in search and synthesis work last year alone, while back-office output increased 10% with 25% fewer people. Client-facing roles grew by 25%, while research analyst, data processor, and administrative support positions shrank by the same proportion.
Thus, McKinsey is not arguing against AI productivity in the abstract; it is arguing that most companies are not capturing the gains McKinsey itself has captured because they are not redesigning workflows. This positions McKinsey as both the most credible voice on the AI productivity paradox and a firm whose consulting work depends on selling solutions to that paradox.
Recommendations
The report offers three recommendations for executives: assess how AI will reshape industry profit pools; build or strengthen AI-powered competitive moats; and turn speed into a structural advantage. These map closely to the multi-year transformation engagements McKinsey sells. The report cites JPMorgan Chase's real-time AI fraud detection, BMW's computer vision quality inspection, and Siemens' AI-coordinated predictive maintenance as examples of current work-acceleration tier, contrasting them with deeper process redesigns that move companies beyond what McKinsey calls the 'gen AI paradox'.
For corporate boards approving AI infrastructure spending, the distinction is the entire investment thesis. Whether the AI capital cycle resembles the long-tail value creation of railroads and electricity, or the wipeout of late-1990s telecom fiber, depends on which side of McKinsey's paradox a given company ends up on.
Source: TNW | Insights News