Forrester: 76% of AI Budgets Still Justified on Productivity, Not the Transformation Labs Are Selling

Forrester analyst Kevin Ogunsua’s read on enterprise AI: frontier labs are still selling existential-risk narratives and transformative futures while 76% of AI decision-makers are justifying their AI investments using plain productivity metrics, with most organizations still stuck in pilot mode or narrowly chasing cost reduction. The disconnect isn’t philosophical — Ogunsua names the actual blockers as governance gaps, unpredictable AI spending, fragmented data estates, unready workforces, and the persistent difficulty of proving measurable business value. None of that gets solved by a lab’s safety roadmap.

That gap between vendor narrative and buyer reality is exactly what Forrester’s own security analysts flagged the day before, arguing CISOs have lost the fight over whether AI works and should instead be forcing a harder conversation about what trustworthy AI actually costs — the security, governance, and workforce investment that most AI business cases quietly leave out. Read together, the two pieces describe the same enterprise buyer: one who has stopped asking “does this AI work” and started asking who pays for the security review, who governs an autonomous agent’s access to credentials, and who trains the junior staff automation is displacing.

Ogunsua’s five recommendations for enterprise leaders are the practical version of that shift — build for value at scale, size governance controls to actual risk, formalize deployment policy, demand vendor transparency on safety testing, and judge initiatives by measured outcomes instead of vendor predictions. For a rep or RevOps leader selling into this buyer, the message is the same one landing across enterprise AI purchasing broadly: trust now has to be earned with evidence and accountability, not with a bigger claim about what the model can theoretically do.