Intelligence from senior operators: de-identified, benchmarked, and decision-ready. Not what vendors sell or analysts predict.
What 610 senior operators have told us across our 6 ZAI surveys, and counting:
senior leaders are very confident in their AI governance.
are running AI agents, but only 40 have a formal oversight policy.
say AI strategy is the hardest skill to find, far ahead of the rest.
Browse by topic · explore the topic pages →
The flagship. Cross-source synthesis of every interview, board, and survey into ranked signals with prevalence and trend.
Drawing on roughly 3,644 senior executives across 188 sessions in August 2026, plus 610 survey responses, this report maps ROI attribution, hidden costs, governance, workforce, and data readiness.
Across the corpus, the binding constraint on AI value is not model capability but the inability to prove returns, control costs, and govern use, leaving budgets resting on intuition.
Findings from roughly 2,119 senior executives across 127 sessions in July 2026, plus 583 survey responses, on governance, workforce erosion, shadow AI, and unproven returns.
The binding constraint on AI value is no longer the technology but the human layer, and the quiet erosion of judgment, trust, and oversight is outpacing the controls meant to catch it.
Drawn from roughly 910 senior executives across 70 sessions in June 2026, plus 444 survey responses, on measurement, governance, workforce, and the cost of compute.
Operators can measure AI activity but not financial return, and they are funding programs on intuition while governance and workforce capability lag far behind deployment.
Quantified signal from monthly operator surveys, benchmarked and dated.
Senior operators are deploying AI cost controls faster than they can see, track, or forecast the spend those controls are meant to govern.
What senior operators said behind closed doors, de-identified and made decision-ready.
Reflections from a finance sector leader on AI hype, the environmental and regulatory cost of compute, and the demand for guardrails.
An operator questioned whether AI is being oversold as new, while pressing for guardrails and attention to its environmental and moral cost.
Insights from a senior operator in enterprise software on human oversight, provenance labeling, spend controls, and deliberate AI rollout.
Experienced operators treat AI as a supervised assistant, governed by human oversight, provenance labeling, spend controls, and deliberate small-scale rollout.
Insights from operators in enterprise software on human-in-the-loop controls, output labeling, board-level spend approval, and cautious adoption.
Operators are turning AI oversight into concrete controls: output labeling, board-approved spend limits, and small supervised pilots with humans accountable for accuracy.
Insights from a nonprofit operator on AI measurement, adoption maturity, and guidance for low-resource, budget-constrained organizations.
Operators cannot easily justify AI spend because productivity gains do not reliably become profit, and guidance ignores resource-constrained organizations.
Insights from an enterprise software advisor and peer group on AI disclosure, spend controls, human oversight, and deliberate rollout.
Operators are treating AI trust as a set of concrete controls: labeling output, tracking spend, and keeping humans reviewing narrow pilots.
Insights from operators across industry on token costs, model selection, governance, vendor lock-in, output evaluation, and strategy-first adoption.
Operators are moving from open-ended AI enthusiasm toward disciplined cost control, strategy-first adoption, and hard scrutiny of whether outputs are actually correct.