Corporate budgets for generative AI tools hit record levels in 2025 while measurable returns stayed flat across most deployments. Axios documented sticker shock at firms facing multimillion-dollar annual bills with limited proof of productivity gains. Primary data from the article shows repeated enterprise complaints over inference costs and integration overhead. McKinsey’s 2025 AI survey of 1,500 executives found only 22 percent reporting significant bottom-line impact from generative tools, down from prior-year projections. Gartner’s peer benchmarks similarly flagged 68 percent of production AI projects as cost centers rather than profit drivers. Coverage overlooked how procurement teams now require explicit payback timelines before approving further spend, a shift visible in Q1 2026 earnings transcripts from Salesforce and ServiceNow. Patterns from 2023-2024 cloud overcommitments repeat here, with AI line items now triggering the same internal audits.