Consumption-Based Pricing Overruns Remain the Dominant Budget Risk for AI Deployments

The Legal Stack's Legal AI Pricing Renegotiation Report 2026 (June) documents that 71% of firms on consumption-based legal AI contracts experienced actual costs exceeding initial projections, with a median overrun of 34% above year-one estimates. Open-ended research and drafting tools showed the highest overruns; document AI and contract review tools were more moderate at 18–22% over projections, largely because document volumes are more predictable than user behavior on open-ended platforms. For mid-sized firm IT directors and COOs managing AI budgets, this data points to a structural risk in any platform that does not offer a credible cost-modeling framework before contract signing. Usage caps, model routing controls, and spending dashboards have moved from nice-to-have features to selection criteria — and vendors that cannot demonstrate a governance path to predictable AI spend are carrying material churn risk as firms hit their first renewal cycle with actual versus budgeted cost comparisons in hand.