Everyone has an opinion on where AI is headed. Fewer people can tell you what’s actually happening inside real companies right now. The current AI trends worth paying attention to aren’t about which model is smartest this quarter. They’re about the growing gap between how much businesses are spending on AI and how little of that spending shows up on the income statement.
That gap is the story of 2026. Adoption has gone mainstream. Returns have not caught up. Understanding why is the difference between running another failed pilot and building something that actually changes how your business operates.
Adoption Is Everywhere. Impact Is Rare.
Enterprise AI use has become the norm rather than the exception. McKinsey’s 2026 State of AI survey found that enterprise AI use in at least one business function sits at 88 percent. Nearly every company you compete with is using AI somewhere in its operations.
But adoption isn’t the same as results. The same research found that only 37 percent of respondents attribute any EBIT impact to AI use at all, a figure that hasn’t moved from the prior year, and just 6 percent qualify as AI “high performers,” organizations attributing meaningful profit impact to AI, which is also flat year over year. Meanwhile, 80 percent of individual AI users report improved personal productivity. People feel more productive. The company’s bottom line doesn’t reflect it. McKinsey enterprise AI ROI 2026: survey data, key stats | Value Add Pulse +2
This is the pattern we see with almost every client who calls us after a disappointing AI rollout. Employees like the tool. Nobody can point to a number that changed because of it. MIT’s Project NANDA put a sharper edge on this problem in its widely cited GenAI Divide research, finding that despite $30 to 40 billion in enterprise investment, 95 percent of generative AI projects yield no measurable business return. The researchers were direct about why: the divide separates the small minority of integrated pilots extracting real value from the vast majority stuck with no measurable P&L impact. Virtualization ReviewVirtualization Review
That’s not a technology failure. It’s a workflow failure. And it’s the single most important current AI trend for any operations leader to understand.
Agentic AI Is Growing Up, Slowly
The next wave everyone’s talking about is agentic AI, systems that don’t just answer questions but take multi-step actions on their own, like processing an invoice from receipt to payment without a human touching every step. Adoption of these tools is real but uneven. McKinsey found that among companies with more than $1 billion in revenue, the share scaling AI agents jumped from 27 percent to 40 percent year over year, a sharp increase. Value Add Pulse
Trust hasn’t caught up to enthusiasm. In McKinsey’s separate research on AI trust maturity, nearly two thirds of respondents cited security and risk concerns as the top barrier to fully scaling agentic AI, well ahead of regulatory uncertainty or technical limitations. That finding matters because it tells you the constraint isn’t imagination or even budget. Companies aren’t struggling to picture what agentic AI could do. They’re struggling to trust it enough to hand over real operational control, and rightly so if the underlying workflow was never designed for an autonomous system to run through it. McKinsey & Company
We think this is actually good news for companies that get the sequencing right. If your competitors are stuck evaluating agentic AI in isolation, waiting for someone else to prove it’s safe, you have room to move first, provided you build the operational foundation before you turn on the automation.
Workforce Anxiety Is Outrunning Workforce Reality
Headcount is another area where perception and outcome have split apart. McKinsey’s 2026 data shows just 14 percent of organizations using AI reported an actual decline in workforce size over the past year, less than half the 32 percent who had expected reductions in the prior year’s survey. Yet expectations for the year ahead climbed anyway, with 39 percent of respondents now anticipating AI-driven headcount decreases, even though two thirds report little or no change in total employment so far.
That’s a pattern worth sitting with. Fear of disruption is consistently running ahead of actual disruption. For leadership teams, that anxiety often becomes a reason to freeze rather than a reason to act deliberately. Neither extreme serves you. The businesses seeing real gains aren’t the ones cutting staff and hoping AI fills the void. They’re the ones redesigning how work moves through the organization so people spend less time on tasks a well-configured system can handle, and more time on judgment calls only a person can make.
Why Workflow Redesign Is the Trend Underneath All the Other Trends
Here’s the thread connecting all of this. Every current AI trend, from the adoption-versus-return gap to agentic AI’s trust problem to workforce anxiety, traces back to the same root cause: companies are bolting AI onto processes that were never built to use it well.
Picture a mid-size property management company that adds an AI chatbot to handle tenant maintenance requests. The chatbot works fine in isolation. But the request still lands in an inbox nobody’s watching closely, gets manually re-entered into a separate work order system, and waits for a human to assign a technician the old way. The AI didn’t fail. The workflow around it did. That’s the exact pattern MIT’s researchers found driving the 95 percent failure rate: tools that don’t fit how the organization already works don’t stick.
This is the core argument behind our BRAVE framework, which starts by mapping how work actually moves through a business before recommending a single piece of AI technology. You don’t fix a workflow by adding a smarter tool to a broken sequence. You fix it by redesigning the sequence first, then choosing tools that fit the redesigned process. Companies that get this right aren’t necessarily using more advanced AI than their competitors. They’re using ordinary AI inside operations that were rebuilt to take advantage of it.
According to Harvard Business Review’s analysis of AI adoption failures, organizations that treat AI as a bolt-on technology purchase consistently underperform those that treat it as an operating model change, because the second group invests in the change management and process redesign the first group skips entirely. That distinction, more than any specific model release or feature update, is the current AI trend that will separate the companies with real returns from the companies still waiting for theirs.
Where This Leaves You
If you’re evaluating your own AI strategy against these trends, ask a blunter question than “which AI tool should we buy.” Ask whether your current workflows could actually absorb a more capable tool if you had one. Most can’t, not because the people running them aren’t capable, but because nobody has redesigned the process around the assumption that AI will be doing part of the work.
That’s exactly the gap we help close. Our team at StrataBlue works with operations leaders to identify where a workflow redesign, not just a new tool, will actually move the numbers that matter. If you want a clear picture of where your own operations stand, get your free diagnostic and we’ll show you exactly where the opportunity is.
