The end of AI blank cheques
The AI spend surge is colliding with financial reality. Cloud is now a major operating cost, forecast variance is rising, and CFOs are taking ownership. AI unit economics are becoming the language that decides which …
Connections
56% of CEOs report no significant financial benefit from AI investments yet [1].
At the same time, AI spending is forecast to reach $2.52 trillion in 2026, up 44% year on year [2].
The blank cheque era is over.
Cloud infrastructure is now the second-largest operating expense at midsize tech companies, behind labour [3]. At SaaS firms, cloud averages around 10 per cent of revenue. AI and ML workloads already represent about 22 per cent of that spend [4].
Traditional FinOps forecasting struggles here.
Training spikes, usage-driven inference, and constant experimentation break the patterns finance relies on. Five to ten per cent monthly forecast variance is now normal [4].
Some signals worth paying attention to:
- 89% of CFOs say rising cloud costs have eroded profitability [4]
- Only 36% feel confident they can drive AI impact [5]
- 97% have formalised cloud governance policies [4]
- Finance involvement doubles forecast predictability versus engineering-only ownership [4]
CFOs are taking direct ownership of cloud spend. The numbers are simply too large to delegate.
For FinOps teams, this is the inflection point.
Tactical wins still matter, but they rarely register in P&L conversations. AI unit economics do. Cost per inference, cost per training run, cost per user query — these translate engineering activity into margins, runway, and stop-or-go decisions.
Without that translation layer, FinOps stays operational while budgets are decided elsewhere. With it, FinOps becomes central to AI investment decisions.
AI budgets are being set now, as part of Q1 planning.
If your FinOps reporting does not connect to executive metrics, you are unlikely to be in the room.
What AI unit economics does your organisation actually track?
Sources
[1] PwC 2026 Global CEO Survey via CFO Dive
https://www.cfodive.com/news/top-5-ai-adoption-challenges-facing-cfos-in-2026/810277/
[2] Gartner AI Spending Forecast
https://www.gartner.com/en/newsroom/press-releases/2025-08-12-gartner-survey-shows-top-priorities-for-cfos-in-2026-include-cost-optimization
[3] CIO.com Cloud Costs Report
https://www.cio.com/article/4110708/cloud-costs-now-no-2-expense-at-midsize-it-companies-behind-labor.html
[4] Cloud Capital – The Cost of Compute
https://www.morningstar.com/news/pr-newswire/20251210ln41991/cfos-take-ownership-as-cloud-becomes-a-financial-risk-factor
[5] Gartner CFO Survey 2026
https://www.gartner.com/en/newsroom/press-releases/2025-08-12-gartner-survey-shows-top-priorities-for-cfos-in-2026-include-cost-optimization
[6] Why most FinOps savings barely register in the boardroom
https://frankcontrepois.com/article/finops-savings-boardroom