A global survey of 13,918 workers found that AI usage climbed while worker confidence in technology fell by 18 percent in a single year. The data describes a workforce using more AI while trusting it less, a pattern that carries direct implications for how leaders think about training, retention, and what it actually means to be AI-ready.
When ManpowerGroup surveyed 13,918 workers across 19 countries for its 2026 Global Talent Barometer, they expected to find the usual story: AI adoption climbing, workers feeling more capable.
They found the opposite.
Regular AI usage jumped 13 percent, reaching 45 percent of the workforce. At the same time, confidence in using technology dropped 18 percent. That was the steepest single-year drop recorded in the report.
The combination of more AI and less confidence carries direct implications for how leaders think about workforce development.
Drop in global worker confidence in using technology in 2026, even as regular AI usage climbed to 45% of the workforce, a 13-point increase year over year.
Source: ManpowerGroup Global Talent Barometer, 2026What "job hugging" actually means for your team
ManpowerGroup named a new pattern in their 2026 data: "job hugging." It describes workers staying put at their current employer out of uncertainty about what AI means for their future rather than out of satisfaction with where they are.
Sixty-four percent of workers surveyed say they plan to stay with their current employer. That sounds like good retention news until you read the next number.
Forty-three percent fear automation may replace their job within the next two years, up five percent from 2025.
So the workforce is staying, but the reason for staying is anxiety rather than commitment. Job hugging is a defensive posture triggered by AI uncertainty that workers have not been given the tools to address.
For a CEO or operations leader, this reads in two ways. First, your most anxious workers are often not your least capable ones. Second, a workforce staying out of fear is not a workforce building new skills. The two things most needed when AI is transforming operations (psychological safety and active upskilling) are exactly what the job-hugging pattern tends to suppress.
Job hugging is a defensive posture triggered by AI uncertainty that workers have not been given the tools to address.
The training disconnect is real, and it is measurable
ManpowerGroup's barometer finds that 56 percent of the global workforce received no recent training, and 57 percent have no access to mentorship opportunities.
Think about that alongside the adoption number. Almost half of workers are now using AI regularly. More than half say they got no training support in the past period. The math on that combination does not work in anyone's favor.
What workers are doing is figuring out AI on their own, finding tools, testing prompts, building workarounds, without a framework for assessing whether what they are doing is actually good. They may be busy with AI tools without getting better at using them. That is the core of what ManpowerGroup calls the AI Confidence Gap: activity without development produces usage without trust.
Of the global workforce received no recent AI training. An additional 57% have no access to mentorship opportunities, even as nearly half of workers now use AI tools regularly.
Source: ManpowerGroup Global Talent Barometer, 2026The training absence shows up by generation, too. The report finds that Baby Boomers experienced a 35 percent drop in tech confidence, and Gen X declined by 25 percent. The two generations that make up the bulk of mid-level management and senior individual contributors are the ones losing confidence fastest.
The pattern here is a development problem wearing technology's clothes.
Why usage and proficiency are not the same thing
When 89 percent of workers say they feel confident about their current role, that sounds like good news. But only 43 percent believe they will still have the same job in two years.
The space between "I'm doing fine today" and "I don't know what next year looks like" is exactly where the confidence problem lives.
Workers are not doubting their current competence. They are doubting their future readiness. And they are not wrong to feel that way if the only preparation they have received is access to a tool.
Access to AI is not the same as knowing how to use AI well. Knowing how to use AI well is not the same as knowing how to think with AI in a way that makes your judgment more valuable, not less. Those are three distinct steps, and many organizations have only cleared the first one.
The 7 Levels of AI Proficiency describes this progression directly. A Level 1 professional (The Cadet, the AI Aware stage) has discovered that AI tools exist and is beginning to experiment. A Level 2 professional (The Ensign, the Prompt Engineer or Practitioner stage) has learned to ask AI good questions. Getting workers from Level 1 to Level 2 requires a framework for improvement they can actually follow, not just more access to tools.
The professionals experiencing the steepest confidence drops are not at Level 0. They are typically somewhere between Level 1 and Level 2, using tools without a clear picture of where they are trying to go. That is a development problem, and the ManpowerGroup data makes the scale of it visible in a way internal surveys rarely surface.
The confidence floor does not hold on its own
The workforce entered the AI era with reasonable confidence. Adoption climbed. Confidence fell. If you extend the trend (higher usage, lower confidence) you get to a place where the workers most exposed to AI are also the least confident about it. That is not a stable condition for any organization trying to build real AI capability.
Leaders who interpret "my team is using AI" as "my team is developing AI capability" are going to find out the hard way that those are different things. The ManpowerGroup barometer puts a number on exactly how different: 45 percent adoption, 18 percent confidence drop. One went up; the other went down.
The resolution is not to slow adoption. The resolution is to build development infrastructure alongside adoption infrastructure so the two curves move in the same direction. That means training that goes beyond tool access, mentorship that connects AI usage to role-specific judgment, and a shared vocabulary for talking about what "getting better at AI" actually looks like inside your organization.
Without that infrastructure, the job-hugging pattern accelerates. Workers stay out of fear rather than growth. The ones who are most AI-ready (your Level 3 and Level 4 professionals, the Critical Thinkers and Context Engineers) will be the first to move when better opportunities appear. The ones who feel most left behind will stay. That is the inverse of the talent distribution most organizations need going forward.
What this looks like in Indiana's workforce
Indiana's economy leans on manufacturing, logistics, and mid-market professional services. Those are the exact sectors where the confidence collapse shows up most visibly in ManpowerGroup's generational breakdown.
The workforce that operates Indiana's production floors, supply chains, and business operations skews toward the generations reporting the steepest confidence drops. Baby Boomers and Gen X account for a substantial share of the experienced, high-judgment workers that Indiana employers have spent years developing. Those workers are now encountering AI tools without a clear development path, and the data says their confidence is eroding as a result.
The IN AI Initiative, announced by Indiana's government earlier this year, positions the state as a leader in AI adoption. What the ManpowerGroup data adds to that picture is the workforce readiness question: adoption infrastructure without confidence development does not produce the workforce outcomes that initiative is designed to create.
For Indiana CEOs reading this, the practical question is direct: does your workforce know where it stands on AI proficiency? The usage number is likely 40 percent or higher if you are paying attention. The question is whether the people using those tools have a shared standard for improvement, a way to measure progress, and a development path that connects their daily AI activity to a visible next level of capability.
That is what moves the confidence number in the right direction.
Related reading: Level 2: The Ensign (Prompt Engineer / Practitioner), the stage where many workers stall when development support is absent.
Three things to do this week
Run a two-question informal survey with your leadership team
Ask two questions in your next team meeting: "How often are you using AI tools at work?" and "How confident are you in your AI skills compared to six months ago?" The divergence between those two answers at your organization is your readiness signal. You do not need a formal assessment to start a conversation. You do need to start the conversation.
Separate "tool access" from "development" in your AI budget
Organizations that have spent on access (licenses, subscriptions, pilots) often have not matched that investment in development (frameworks, training that builds on itself, a shared vocabulary for progress). Look at your AI spend this quarter. If the development line is zero, the confidence data predicts what will happen to your workforce confidence next quarter.
Give your team a proficiency baseline, not just a tool
The 7 Levels of AI Proficiency assessment at assess.launchready.ai takes under ten minutes and places individuals across seven stages of AI capability. It is free. It gives workers a specific, named stage to start from and a clear picture of what the next level requires. A worker who knows they are at Level 2 and understands what Level 3 demands is in a fundamentally different position than a worker who uses AI every day with no standard for self-assessment.
Sources
- ManpowerGroup. "Global Talent Barometer 2026: AI Use Accelerates as Worker Confidence Falls and 'Job Hugging' Takes Hold." January 2026.
- ManpowerGroup. Global Talent Barometer 2026 Global Report PDF. 13,918 workers, 19 countries, fieldwork September to October 2025.
- Fortune. "AI adoption is accelerating, but confidence is collapsing." January 21, 2026.
- LaunchReady.ai. "What Is AI Proficiency: A Complete Guide for 2026."
- LaunchReady.ai. "Indiana IN AI Initiative: What It Means for CEOs."
Frequently Asked Questions
What is the ManpowerGroup Global Talent Barometer 2026?
The ManpowerGroup Global Talent Barometer is an annual survey measuring workforce sentiment, confidence, and employment trends. The 2026 edition surveyed 13,918 workers across 19 countries, with fieldwork conducted from September through October 2025. It tracks worker confidence across technology use, job security, and skills readiness. The 2026 report was the first to measure AI usage as a distinct metric alongside technology confidence.
Why is AI confidence falling even as adoption rises?
ManpowerGroup's 2026 data shows that more workers are using AI tools regularly (45 percent, up 13 percent from the prior year), yet technology confidence fell 18 percent. That was the steepest single-year drop recorded in the report. The most likely explanation is that workers are gaining access to AI tools without gaining the development support needed to use those tools well. Usage without a framework for improvement produces activity, not confidence. Training shortfalls compound this: 56 percent of the global workforce received no recent training, and 57 percent have no access to mentorship.
What is job hugging and why does it matter?
Job hugging is the term ManpowerGroup uses to describe workers who plan to stay with their current employer out of uncertainty and anxiety about AI rather than out of positive commitment. Sixty-four percent of workers surveyed plan to stay, while 43 percent fear automation may replace their job within two years. Job hugging is a defensive posture, not a loyalty signal, with direct implications for productivity, morale, and AI capability development over time.
Which generations are most affected by the AI confidence collapse?
According to ManpowerGroup's 2026 barometer, Baby Boomers reported a 35 percent drop in tech confidence and Gen X declined by 25 percent. These generations make up a large share of experienced mid-level managers and senior individual contributors at most organizations. They are the people most likely to be managing AI-tool adoption decisions, not just using the tools themselves. The decline is not evenly distributed, and the steepest drops are among the workers whose judgment is most consequential to organizational outcomes.
What is the 7 Levels of AI Proficiency and how does it help address the confidence collapse?
The 7 Levels of AI Proficiency is a framework developed by Harrison Painter at LaunchReady.ai that describes seven stages of AI capability: Level 1 (The Cadet, AI Aware), Level 2 (The Ensign, Prompt Engineer or Practitioner), Level 3 (The Lieutenant, Critical Thinker), Level 4 (The Commander, Context Engineer or Builder), Level 5 (The Captain, Design Thinker), Level 6 (The Admiral, Systems Integrator or Leader), and Level 7 (The Mission Director, AI Orchestrator). The framework gives workers a shared vocabulary for AI development: a way to name where they are and measure progress over time.
What can a CEO do right now to address the AI confidence collapse?
Three actions carry the most weight. First, survey your leadership team informally on the divergence between their AI usage frequency and their AI confidence. Second, audit your AI budget for the split between access spending and development spending. Many organizations have invested heavily in access and minimally in development. Third, give your team a proficiency baseline through the free 7 Levels of AI Proficiency assessment at assess.launchready.ai, which places individuals in under ten minutes and gives them a specific stage and a clear path forward.
How does Indiana's workforce connect to the ManpowerGroup findings?
Indiana's economy is weighted toward manufacturing, logistics, and mid-market professional services. Those are the sectors that employ a large share of Baby Boomer and Gen X workers experiencing the steepest confidence drops in ManpowerGroup's data. The state's IN AI Initiative creates adoption pressure. The ManpowerGroup barometer adds the workforce readiness dimension that adoption pressure alone does not address: 45 percent of workers using AI tools, while confidence falls and training support is absent for the majority.
What is the difference between AI adoption and AI proficiency?
AI adoption measures whether workers have access to AI tools and are using them. AI proficiency measures whether workers can use AI in ways that improve their judgment, output quality, and decision-making over time. ManpowerGroup's 2026 data shows that adoption and confidence can move in opposite directions. That is exactly what happens when access outpaces development. Proficiency requires a framework for progression, structured development support, and a way to measure improvement at both the individual and organizational level.
Find your AI Proficiency level
The free 7 Levels assessment places you across seven stages of AI capability. Under ten minutes. Research-backed scoring.