McKinsey's Quantum Black research arm published the most widely cited economic estimate of generative AI's potential: between $2.6 and $4.4 trillion in annual incremental value across the global economy, with banking ($200–340 billion), retail ($400–660 billion), and high-tech ($240–460 billion) the highest-impact sectors. The estimate is built bottom-up across 63 specific use cases rather than top-down by industry — making it more defensible than the breathless valuation projections that have crowded the space since.
Four functions account for roughly three-quarters of the modeled value: **customer operations** (AI agents handling Tier-1 support), **marketing and sales** (personalized content at scale), **software engineering** (code generation and review), and **R&D** (drug discovery, materials science, design exploration). The report's most important caveat is implementation — most organizations capture only a fraction of the theoretical value because production deployment requires data infrastructure, change management, and trust-building that few have yet built well.
For AI Academy fellows building startups or advising corporations, this report is foundational reading. The numbers themselves matter less than the bottom-up methodology: a credible way to argue for AI investment to a CFO or board, sector by sector and function by function, rather than via hype. The same framework underpins the Investment Alliance's screening criteria for portfolio companies.
Source: McKinsey & Company. Read the full report at <https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-economic-potential-of-generative-ai-the-next-productivity-frontier>.