The AI investment conversation is changing. T&I News recently captured that shift as businesses began asking harder questions about whether rapid spending on artificial intelligence is producing the performance promised. The latest evidence suggests a useful answer: many organizations are measuring the wrong unit of transformation.
McKinsey’s State of AI in 2026 found a striking gap between AI high performers and everyone else. Nearly three-quarters of high performers report fundamentally redesigning workflows because of AI use, compared with about one-quarter of other respondents. High performers are also more likely to have defined processes for measuring impact.
That matters for the Gulf. The UAE and its neighbors have invested heavily in AI infrastructure, platforms, talent, and public-sector transformation. The next competitive question concerns how quickly institutions turn those assets into measurable operating results. Buying a stronger model or another enterprise license will rarely answer that question on its own. Leaders need to redesign how work moves from request to decision to execution.
Start With the Outcome, Not the Tool
The common pattern is easy to recognize. A company buys an AI assistant, gives employees access, runs training, and then counts logins or prompts as evidence of adoption. Employees use the tool to draft a document, summarize a meeting, or speed up a familiar task. Some time gets saved, but the broader process remains unchanged. The organization has made one step faster while preserving every old handoff, approval, queue, and duplication around it.
The stronger approach starts with a business outcome. Reduce loan-processing time. Shorten customer-response cycles. Improve the accuracy of compliance review. Increase conversion in a sales workflow. Once the outcome is clear, leaders can map the full process and ask where AI should assist, where it should act, where a person must decide, and where a handoff can disappear entirely.
This is why McKinsey’s July research on AI transformation is so important. Leaders were 5.3 times more likely to report enterprise value capture when workflows were redesigned than when they remained unchanged. The finding points away from tool deployment as the goal and toward work redesign as the real source of value.
Define the Human Job Before Automating the Machine Job
Workflow redesign also forces a question that many AI programs postpone: what exactly should people do after AI changes the task? If an agent can prepare the first analysis in minutes, who validates it? Which decisions still require human judgment? What happens when the system is uncertain? What should employees do with the time that automation releases?
Those questions sound operational, but they are also psychological. Employees are more likely to engage seriously with AI when they understand how their expertise remains useful and how the organization will use the capacity they free up. If leaders talk about efficiency while staying vague about roles, people reasonably wonder whether every productivity gain will become a head-count target.
A good redesign therefore assigns both machine responsibilities and human responsibilities. The AI may draft, classify, retrieve, compare, or recommend. The person may set the objective, supply context, judge exceptions, verify critical claims, make high-stakes decisions, and remain accountable for the result. That clarity improves quality and reduces resistance at the same time.
Measure the Workflow, Not the Excitement
The next mistake is measurement. Usage matters, but it only tells leaders whether people touched the tool. It does not tell them whether the organization created value. A useful AI scorecard should connect at least four dimensions: efficiency, quality, business impact, and employee experience.
For efficiency, measure cycle time, cost per transaction, or hours saved against a baseline. For quality, track error rates, rework, escalation, or customer satisfaction. For business impact, connect the workflow to revenue, retention, faster time to market, or another strategic outcome. For employee experience, ask whether AI is actually removing low-value work and whether people feel confident using it responsibly.
Microsoft’s 2026 Work Trend Index reinforces the organizational point. Its global analysis found that culture, manager support, and talent practices accounted for more than twice the reported AI impact of individual factors such as mindset and behavior. In other words, capable employees still struggle when the system around them rewards the old way of working.
Make Workflow Redesign the Unit of AI Transformation
For Gulf organizations, the practical implication is straightforward. Each major AI investment should be attached to a named workflow, a baseline, a target business outcome, a clear human-machine division of labor, and a small set of metrics. Leaders should scale only after the workflow performs better in practice, not because the demonstration looked impressive or the vendor reached an implementation milestone.
The region does not need to slow its AI ambition. It needs to aim that ambition at the operating model. The organizations that pull ahead will treat AI as a reason to redesign work, clarify accountability, and measure real outcomes. That is how AI spending turns into AI return.
