Accenture named the dynamic explicitly. The consulting giant cited "AI demand compression" when cutting its forward guidance — the first major firm to use that phrase directly.1
The pattern points to a ROI credibility problem. Enterprise clients adopted AI tools through 2024 and 2025. Now they are pausing spend pending measurable returns. Vendors who built growth forecasts on AI upsell cycles are repricing.
The damage is concentrated in the software and consulting layer — companies that sell AI to enterprises, not companies that build AI infrastructure. The infrastructure layer, chips and cloud compute, has held relatively firm. The application layer has not.
Three signals define the moment. The declines are clustered: Salesforce, Adobe, and Atlassian are not direct competitors, yet all are falling simultaneously.1 Accenture's guidance language targeted demand, not macro conditions or budget cycles. And the declines have accelerated beyond what sector rotation or interest rate sensitivity explains.
Near-term pressure is expected to persist for one to two quarters as clients wait for clearer ROI evidence before resuming spend.1 IBM, Infosys, and Cognizant — each with material AI consulting revenue — face the same compression dynamic.1
The market is re-rating enterprise AI as a slower-growth category than previously priced. Multiples built on 2024-era adoption projections are contracting. Vendors with diversified legacy revenue streams may absorb the adjustment. Pure-play enterprise AI software faces the harder path.
The core investor question: demand pause or structural ceiling? A pause implies recovery once enterprise clients complete ROI assessments. A ceiling implies the enterprise AI services market is smaller than consensus assumed. The synchronized nature of the current underperformance suggests the ceiling risk is underpriced.
