Employers Say Training Works. Their Own Workers Disagree.

Four signals this week share one thread: the gap between what employers believe about workforce readiness and what the data actually shows. A frontline-industry survey finds employers grading their own training too generously. An enterprise AI-readiness study finds barely a tenth of employees can be trusted to check an AI system's output. Federal money is moving to close the AI-skills gap specifically. And a drone manufacturer just opened another practical, no-degree route for veterans into a new trade. Read together, they describe a market where the willingness to invest in skills is real, but the instruments to verify what that investment actually produced are still catching up.

Frontline employers grade their own training too generously

A survey conducted by Chegg and covered by HR Dive polled 1,000 employers and 1,005 workers across 10 frontline-heavy US industries, including manufacturing, retail, IT, and finance. The headline finding: employers name a lack of practical skills as their number-one workforce problem, and more than half say entry-level workers arrive unready for the job. Nearly a third of employers report spending the equivalent of a full workday compensating for skills gaps in their teams — in manufacturing specifically, that figure rises to 46%.

The more revealing number is the perception gap. 77% of employers rate their own training programs as effective. Only 58% of employees agree. Roughly 40% of workers say they don't get enough hands-on practice, and 51% describe their training as too generic, disconnected from the actual tasks they're asked to perform.

This is the gap PowerTechs is built to close. Employers aren't wrong that skills gaps are costly — they're wrong about how well their current training closes them. A workforce that rates its own training effective while the people going through it say otherwise is a workforce running on a training loop that never checks its own output. Practical demonstration, not a completed course, is what would settle the disagreement. (Source: https://www.hrdive.com/news/training-gaps-in-frontline-heavy-industries-thwarting-workforce-readiness/823243/)

13% of employees can actually verify AI's work

Workera, drawing on a new IDC report, puts the global cost of the skills gap at $5.5 trillion by 2026 — lost product timelines, quality problems, and lost revenue as more than 90% of organizations run into critical skill shortages. But the more specific number comes from Workera's own assessment base of over 88,000 evaluations: only 13% of enterprise employees currently hold the critical skills needed to work with AI agents. The other 87% lack the task-proficiency and judgment to reliably verify, correct, or redirect what an AI system produces in production. Fewer than a third of organizations consider themselves fully ready for AI-driven work.

That 13% figure matters more than the $5.5 trillion headline. A dollar figure is easy to file away as an abstraction. A statement that 87% of the workforce cannot be trusted to catch an AI system's mistakes is a concrete, testable claim about a specific skill — one that a resume or a completed AI course cannot answer, and a performance-based assessment can. (Source: https://www.workera.ai/blog/the-5-5-trillion-skills-gap-what-idcs-new-report-reveals-about-ai-workforce-readiness)

$25 million for AI upskilling, apprenticeships next

The Commerce Department's Economic Development Administration opened a Notice of Funding Opportunity for its AI Upskilling Accelerator Pilot Program: $25 million to fund the design and rollout of AI-skills training, aimed at emerging and critical industries, with applications accepted through July 10, 2026. In parallel, the Department of Labor is moving to build AI-skills content directly into Registered Apprenticeships nationally, rather than leaving AI training as a separate, optional add-on.

Neither program is large on its own. What matters is where the money is aimed: practice-based training structures — apprenticeships, accelerator pilots — rather than standalone coursework. It's a small but concrete extension of the same pattern showing up across Build Freedom, VET TEC 2.0, and Meta's America's Workforce Academy: federal and philanthropic capital increasingly prefers earn-and-learn mechanisms over classroom credentials, now extending that preference specifically to AI skills. (Source: https://www.eda.gov/news/press-release/2026/05/11/us-department-commerce-announces-25-million-notice-funding)

A drone maker backs veteran pilots into a new trade

Lucid Bots, the Charlotte-based maker of the Sherpa drone, announced it has become an official sponsor of Vets to Drones, a veteran-led nonprofit founded in 2023 by Marine Corps veteran Chris Lewis. The organization moves transitioning service members into commercial drone careers through free preparation for FAA Part 107 certification, plus advanced tracks in mapping, public safety, infrastructure inspection, and agriculture. Training follows NIST flight standards, backed by mentorship and job placement, and the program has already supported thousands of veterans nationally.

This is a smaller signal than a federal program, but it points at the same structure: a specific employer, a specific credential (FAA Part 107), and a specific placement pipeline, aimed at veterans entering an emerging manufacturing and autonomous-systems field. It sits alongside the growing cluster of earn-and-credential routes — data centers, semiconductors, energy — that increasingly treat military transition as a direct pipeline into critical industries, not a side program. (Source: https://www.prnewswire.com/news-releases/lucid-bots-partners-with-vets-to-drones-to-open-commercial-drone-careers-for-transitioning-veterans-302819582.html)

Takeaway

Every signal this week points at the same unresolved question: employers believe they're investing in the right training, but the data — from frontline workers who disagree with their own employer's assessment, to an AI-readiness study that quantifies just how few employees can be trusted to check a machine's work — says the verification layer is still missing. Federal money and corporate sponsorships are expanding the on-ramps into critical industries. What still needs building is the instrument that confirms, in a format an employer can trust, that the training actually produced the skill it promised.

For veterans: see roles that match your military experience

For employers: book a 30-minute call