Your Team Is Using AI Every Day, But Most of Them Aren’t Getting Any Better at Their Jobs Because of It
Sign Up For Daily Job Market News
Every company I talk to right now is handing out AI tools fast, but getting real value from them is proving to be a completely different problem.
A new Q2 2026 report from NROC Security USA looked at how employees actually use generative AI at work, based on 4,800 business users and roughly 139,000 GenAI interactions. Active usage jumped to 31% of eligible employees, up from 19% in the first quarter. The number of AI app families in use nearly tripled, from 19 to 53.
Here’s the gap that matters more than the adoption numbers: only about 5% of employees using these apps qualify as truly effective users, the ones actually getting meaningful productivity gains.
It’s a tenfold jump from Q1, real progress by any measure. It also means 95% of the people using these tools aren’t seeing much from them.
The reason behind that gap matters more than the number itself. NROC’s data shows that frequency of use and prompting skill drive productivity, not simple access to the tools. A company can buy enterprise licenses for every employee and still see almost no gain if people open the tool occasionally and keep their prompts basic.
“Frequency and prompting skill, not AI access, drive productivity gains,” NROC CEO Antti Reijonen said. He put the core tension even more bluntly: “We encourage GenAI use, but are businesses becoming more productive? It’s the hardest question to answer.”
I think most companies are measuring AI the wrong way… They’re counting seats and logins. This report says that what actually moves the needle is frequency and prompting skill, and only 5% of users currently have both.
Put that next to the productivity numbers economists are watching, which I get into below. National productivity growth is stuck near 1%, even with AI tools everywhere. This study explains part of why.
The tools are in everyone’s hands, but the skill to use them well isn’t, so the output gains show up for a small slice of users and disappear into the average for everyone else.
For anyone hiring right now, that reframes what “AI skills” means on a resume. Almost every candidate has opened ChatGPT by now, so that alone tells you nothing. The real test is whether they use it daily and consistently get better results out of it than the person next to them.
We’ve started weighing that distinction on job orders where clients ask us to screen for AI fluency, and it’s a real, testable difference. It’s about to matter as much as any software proficiency we screen for today. If you’re building out a role where AI use is part of the job, our team can help you write a screening process that actually tests for it rather than just asking candidates to check a box- the kind of thing we build into a search when you work with our staffing team.
The practical fix here is cheaper than most employers expect. Training people to prompt well and building the habit of reaching for the tool in the first place moves the needle more than another round of licenses. It’s a coaching problem, and coaching costs a lot less than the software you’re already paying for.
Staffing your team doesn’t have to be hard.
Reach out and see how we can help.
AI’s Payoff Depends on the Paychecks It Might Cut
Joseph Carson, a longtime economist who spent years as Director of Global Economic Research at AllianceBernstein, makes a simple argument in a new Haver Analytics piece with big implications for anyone planning headcount around AI. AI can only pay off economically if the people it might replace keep their jobs and their paychecks.
Here’s the logic: consumer spending drives roughly 70% of U.S. GDP growth, with housing adding another 3%, while business investment accounts for less than 15%.
Consumer spending only holds up when people have jobs and income. The whole engine runs on employment.
Every major past innovation has created more jobs than it destroyed, and Carson’s own view is direct: “All past innovations have ultimately created more jobs than they eliminated.” His worry is that AI is built differently, designed to do tasks people currently get paid for.
“AI is designed to perform tasks traditionally done by humans, whereas the Industrial Revolution was a massive job creator,” Carson said.
The first half of 2026 shows why the timing matters. GDP grew at a 1.75% annual rate, consumer spending rose 1.8%, and the economy added 450,000 payroll jobs, roughly 75,000 a month, a soft pace by recent standards. Productivity grew only about 1%, so whatever AI productivity boom is coming isn’t showing up in the national numbers yet.
Carson’s math is the part that should give any leader pause. “If real consumer spending decreases by 100 basis points due to sluggish job growth, investments in AI models and infrastructure would need to increase by 3X or 4X above the current rate to maintain the same GDP growth rate,” he said. Business investment is too small a slice of the economy to replace lost consumer demand if job growth stalls.
The catch is circular. Carson put it plainly: “If AI enables companies to downsize their workforce or eliminates the necessity to expand it, this could significantly harm overall economic performance.” If AI lets companies shrink their workforces, it can undercut the same consumer spending that AI vendors ultimately need to sell into.
Most of the AI jobs debate gets stuck on one question: will the robots take my job? Carson is asking the bigger one… If they do, who’s left to buy anything?
The number I keep coming back to is 70%. Consumer spending is roughly 70% of GDP growth, and consumer spending runs on employment.
You can automate a call center or a coding team and book the savings this quarter. Do it across the whole economy, and you’ve thinned out the paychecks that were buying the products those companies sell.
I’d push back gently on one part of this. Every prior wave of technology also looked like a pure job killer while it was happening, and the new jobs showed up on a delay nobody could predict in advance. Carson admits as much himself.
The honest read is that we don’t know yet, and the first half of the year’s data doesn’t settle it either way. What I’d tell any leader planning big AI-driven headcount cuts: the savings are real and immediate, the demand risk is real and slower to show up, and you’re betting on both at once.
If you’re weighing that decision for your own team, our staffing services give you the flexibility to scale without making an irreversible cut you might regret in a year.
Employees Don’t Rate Their CEO on Charisma, But Whether Layoffs Feel Likely
Glassdoor dug into what employees actually say about the CEOs who win its Best CEO award, and the answer comes down to one thing: they help people grow their careers.
At companies with a Best CEO, employees mention internal mobility 253% more often than at other companies, career development 166% more, growth opportunities 159% more, and career growth 111% more, according to Glassdoor’s analysis of more than 150,000 reviews from U.S.-based, non-intern employees submitted between May 16, 2025 and May 15, 2026. The specific practices that build those careers show up in the data too: coaching mentions run 149% higher, promoting from within 142% higher, training 120% higher, and mentorship 105% higher. Culture is the other big theme; reviews at winning companies mention “culture” 85% more, “inclusive” 82% more, and “work-life balance” 38% more.
Now flip it. At companies whose CEOs didn’t win, the loudest complaint by far is job security. “Layoffs” comes up 304% more often than at winning companies, “turnover” 130% more, and “fear” 91% more.
Right behind that sit trust and respect problems: “respect” mentions run 162% higher, “trust” 151% higher, and “toxic” 111% higher. Employees at these companies also describe a leadership team with no clear direction: “strategy” appears 154% more often, “direction” 104% more, and “alignment” 99% more. Communication gaps round it out, with “transparency” up 109% and “communication” up 76%.
Glassdoor’s chief economist, Daniel Zhao, found something else that stood out: the best CEOs usually aren’t the larger-than-life personalities employees rave about. His read is that competent leadership builds “a culture competently meeting the baseline expectations of employees, while also creating opportunities for growth,” and stays relatively unremarkable to employees day to day. On CEO visibility, no news is mostly good news.
The headline number everyone will quote here is internal mobility, up 253% at the best-run companies. Pay attention to the flip side too, because it’s the more useful signal for anyone hiring or running a team right now: at the worst-rated companies, “layoffs” shows up 304% more than anywhere else.
Those two facts are the same story told from opposite ends. When people can see a path forward inside a company, they talk about growth. When they can’t, they talk about fear. Retention lives or dies on which of those two conversations is happening inside your building.
Career development is the signal employees use to decide whether they’re building something where they are or just waiting for the next round of cuts. Promoting from within and real coaching cost far less than the turnover they prevent, and they’re the clearest signal a company can send that it plans to keep the people it hires. If your team wants candidates who’ll stick around and grow instead of churning through another round of quick hires, we connect employers with candidates who fit that longer view.
