AI payoff remains elusive outside tech

Data through October 2026 shows a widening divide between technology giants and mainstream corporate America. Revenue per worker and profitability are rising among the Magnificent Seven, while revenue per worker and profit margins remain flat across the S&P 493. Total Factor Productivity data adjusted by the San Francisco Fed for cyclical changes in factor utilization also shows no evidence of an AI-driven acceleration in output growth.

The lack of measurable returns is shaping spending decisions. Mainstream executives increasingly view AI token consumption as a cost center rather than a revenue engine. Venture-backed startups continue expanding their usage, but mainstream businesses are unwilling to pay premium prices for the latest models. That divergence limits the broader economic impact while agentic AI and heavy token consumption remain concentrated in a narrow segment of the economy.

Lower model prices alone do not resolve adoption barriers. Business decisions depend on context, accountability, trust, cultural expectations and ethical trade-offs. These judgment requirements affect reputational risk, revenue growth and margins, leaving mainstream companies with challenges beyond the cost of deploying AI.