A study published by the Boston Consulting Group in June 2026 surfaces a number that should be in every boardroom discussion about AI investment: 74% of frontline workers already use artificial intelligence regularly — up 23 percentage points from 2025. Among frequent users, 42% say they save at least a full day of work per week.
So far, it’s a success story. The problem shows up in the next line of the research: 66% of these professionals receive little to no guidance on what to do with the time they save, and more than half say it isn’t being redirected to higher-value work. The company bought the tool, the tool worked, and the gain is evaporating — without anyone having decided that it should.
The decision moved up a level. The problem didn’t move with it
That number looks stranger next to another survey from the same firm, from January 2026: 72% of CEOs now identify themselves as the primary AI decision-maker inside their own company — double the year before. The stated logic is that AI decisions affect revenue, customer experience, and the structure of work all at once, and no single functional executive has the mandate to decide across all those fronts alone.
In other words: the AI decision has already reached the right table. And yet, the time the tool frees up keeps getting lost at the operational level. That points to a bottleneck that isn’t about executive sponsorship — it’s about process redesign.
The mistake behind the data, according to João Paulo Batistella
João Paulo Batistella, an innovation executive and organizational transformation specialist, describes exactly this pattern when explaining why corporate AI projects deliver less than promised: “most are trying to automate the legacy, when they should be using artificial intelligence to redesign the business itself.” A tool that saves time inside a process that hasn’t changed doesn’t create new value — it just makes the old process faster, with the freed-up time left with nowhere defined to go.
Batistella also points to the root cause of why this keeps happening even with the CEO driving the decision: “putting a new tool on top of an old structure doesn’t mean transforming the business.” Approving the investment at the right level is necessary, but not sufficient — what’s missing is the less visible step of deciding, process by process, what changes in the work once the time is freed up.
Three questions before the next AI adoption report
- When a team starts saving time with AI, is someone responsible for deciding where that time should go — or does it just disappear into the routine?
- Is the metric the company tracks “how much time did AI save” or “what new outcome did that time generate”? Those are different questions, and only the second one measures value.
- If the answer to the two questions above doesn’t exist, the AI investment is being treated as a tool — not as process redesign.
The BCG data isn’t a reason to slow down AI adoption — it’s a reason to treat the next step with the same rigor given to approving the investment.
Follow Eleva Tecnologia for more analysis on technology applied to business: follow @ElevaTechno on X or @elevatechnologies on Instagram, or learn more about the group at elevatec.net/about.
