60% Want to Hire. 44% Cannot Fill the Jobs.

The labor market sent what looked like several conflicting signals this week.

Sixty percent of U.S. employers plan to increase headcount during the second half of 2026.

Yet 44% already have positions they cannot fill—the highest level since spring 2023.

At the same time, overall hiring has slowed, permanent recruitment remains cautious, temporary staffing is accelerating, and AI may be creating new demand in some of the occupations it was expected to destroy.

These are not conflicting signals.

They point to the same conclusion:

The problem is no longer simply a shortage of jobs or workers. It is a shortage of deployable capacity.

That distinction matters enormously for staffing.

Employers still need more work done

The strongest number this week came from Express Employment Professionals.

Its Harris Poll survey of 1,006 U.S. hiring decision-makers found that 60% expect to add employees during the remainder of 2026, including 19% planning significant increases.

Why are they hiring?

53% have more work to complete.
49% are creating new positions.
42% must replace people lost through turnover.
37% are expanding into new markets or business categories.

This is real demand—not theoretical optimism.

But 44% of employers currently have roles they cannot fill, up from 36% last fall. Nine in ten expect hiring challenges during the rest of the year.

The staffing opportunity is sitting directly inside that gap.

But filling it will require more than sending additional résumés.

The market is buying flexibility before commitment

The UK offered another important signal.

KPMG and the Recruitment and Employment Confederation reported that temporary billings rose at their fastest rate since April 2023, while permanent staff appointments continued to decline—but at a much slower pace than in May.

Employers are still progressing projects. They are simply becoming more cautious about taking on permanent fixed costs while uncertainty remains high.

That is not just a temporary-staffing recovery.

It reflects a broader change in how companies are thinking about capacity.

They increasingly want:

Capacity that can expand and contract.
Skills that can be deployed quickly.
Outcomes without unnecessary fixed overhead.
Human expertise supported by automation.
Less operational risk when demand is uncertain.

The staffing firms positioned only to sell permanent headcount may struggle.

The firms able to provide flexible human, digital and blended capacity could become much more valuable.

Hiring is slowing—but demand has not disappeared

The headline U.S. employment numbers were weak.

According to SHRM’s review of the latest Bureau of Labor Statistics report, employers added only 57,000 jobs in June. April and May were revised down by a combined 74,000 jobs.

But the weakness was uneven.

Professional and business services added 36,000 jobs. Healthcare added 22,000. Manufacturing and construction also posted gains as investment continued around AI-related infrastructure.

That does not look like a market where nobody needs workers.

It looks like a market where employers are becoming far more selective about which work justifies additional headcount—and how that capacity should be acquired.

Companies still need output.

They are simply asking harder questions before carrying more permanent cost.

That is exactly where staffing should be strongest.

But only when we sell the solution to the work—not merely access to another worker.

AI may be creating jobs differently

Indeed Hiring Lab published one of the most interesting labor-market signals I have seen this year.

Since the introduction of widely available agentic coding tools, U.S. software-development job postings have risen by almost 15%, while overall job postings declined by 7%.

Even more telling:

71% of the increase came from senior roles.
37% came from jobs mentioning AI in the title.

Indeed is careful to say that correlation does not prove causation. So should we.

But its analysis suggests that the relationship between AI exposure and employment growth may be changing. Over the past year, some of the occupations most exposed to AI have also shown some of the strongest rebounds.

This supports something I have been saying for some time:

AI does not only automate work. It changes the economics of attempting more work.

When one skilled developer, recruiter, analyst or operator can achieve substantially more, the business may not simply remove that person.

It may decide to build more products, serve more customers, enter more markets or launch projects that previously did not make economic sense.

But the growth appears concentrated around people who can direct the technology—not people who simply perform the old process manually.

That is the warning for every staffing business.

Demand may remain strong while the specification of the worker changes underneath us.

The new shortage is ready-to-deploy talent

Government and industry are already responding to the capacity problem.

The U.S. Department of Labor has awarded nearly $162 million through five cooperative agreements to expand Registered Apprenticeships in occupations tied to shipbuilding, defense, emerging technology and American reindustrialization.

Jobs for the Future received $40 million to build apprenticeship pipelines for the AI, semiconductor and nuclear-energy infrastructure sectors.

JFF cited projections that the annual shortage of skilled tradespeople could approach half a million over the coming decade.

Employers are also looking internationally and seasonally for capacity. For October 1, 2026 start dates, the Department of Labor received applications covering 51,158 H-2B worker positions—well above the 33,000 statutory allocation for the first half of the fiscal year.

Look at the pattern:

Employers need workers.
They cannot always find workers with the right skills.
They are hesitant to carry excessive permanent cost.
They are investing in training, contingent labor, international talent and automation.

The demand is not going away.

The mechanism for fulfilling it is fragmenting.

Staffing must move from recruiting to capacity design

Eastridge described the recruiter’s evolution this week as a move from transactional recruiting to strategic talent advisory.

Modern recruiters increasingly need business acumen, labor-market data and consultative ability—not sourcing expertise alone.

I agree, but I think the transition goes one step further.

The future staffing partner will help a client decide:

Which work must remain human?
Which work can be completed by a digital worker?
Where should humans and AI operate together?
Which skills can be trained instead of bought?
Which capacity should be permanent, temporary, outsourced or on demand?
How will the business measure output, quality and return?

That is not recruitment administration.

That is workforce architecture.

And it is why removing administration from staffing is so important.

A recruiter spending the day moving information between systems, chasing updates, re-keying data and manually coordinating workflows cannot become a strategic adviser.

Digital workers should absorb that administrative work so human recruiters can spend their time understanding businesses, building relationships, assessing potential and designing better workforce solutions.

What we worked on at EQ.app this week

This week, we mapped 14 real staffing workflows into a clearer operating model—from candidate screening and sourcing to timecards, payroll, billing, collections, onboarding and client-system integration.

The common thread was not “use more AI.”

It was much more practical:

Define the work.
Break it into clear duties.
Assign each duty to a human or digital worker.
Put rules and approvals around the sensitive decisions.
Measure completed, checked work.
Increase capacity only after value is proven.

We also spent time clarifying why digital-worker pricing should not look like traditional software pricing.

You are not buying another seat or another tool for recruiters to manage.

You are funding governed work that gets completed.

That distinction—the move from access to output—is where I believe the broader staffing market is heading too.

My take

This week’s numbers are not telling us that staffing is dying.

They are telling us that the traditional staffing model is becoming incomplete.

60% of employers want to hire.
44% cannot fill existing roles.
Temporary staffing is accelerating.
AI-exposed jobs may be rebounding.
Major investment is flowing into the training pipelines required for new infrastructure and industries.

The opportunity is enormous.

But the winners will not simply be the firms with the largest candidate databases or the most recruiters making calls.

They will be the firms that can assemble the right capacity—human and digital—around the work that needs to be completed.

The question for staffing leaders is no longer:

“How many people can we place?”

It is:

“How much productive capacity can we create for our clients—and how quickly can we deploy it?”

That is the market staffing is entering now.