Welcome to this week’s StaffingHub Brief, your strategic intelligence roundup for staffing agency leaders. In this week’s issue:
- Only 25% of AI initiatives deliver the ROI CEOs expect and just 16% have scaled company-wide, yet almost two-thirds of CEOs still worry they’re underinvesting in AI.
- Global AI infrastructure spending hit $318 billion in 2025, up 81%, and demand for skilled trades and AI-adjacent roles is already outrunning the rest of the job market.
- Courts are holding companies liable for what their outsourced AI vendors get wrong, not the vendor that built the model.
Hiring is growing, but the gap between open roles and filled roles just hit a three-year high
Temp jobs jumped by 9,300 in June, more than economists expected, even as total nonfarm payrolls added just 57,000 against a forecast of 113,000, according to BLS data reported by Staffing Industry Analysts. The unemployment rate slipped to 4.19% from 4.30%, and temp penetration held flat at 1.57%. (Learn more)
Employers still plan to hire, they just can’t find who they need. An Express Employment Professionals-Harris Poll of 1,006 hiring managers found 84% feel positive about the second half of 2026 and 60% plan to grow headcount, driven mainly by rising workloads and newly created roles. But 44% currently have open positions they can’t fill, up from 36% last fall and the highest share since spring 2023. (Learn more)
Why it matters: Demand is real. The fill-rate gap is where a well-positioned firm wins share right now, and it’s won on speed to the candidates everyone else can’t find, not on volume.
The AI ROI everyone’s chasing is running years behind the budget cycle
Only about 25% of AI initiatives deliver the return CEOs expect, and just 16% have scaled across the enterprise, per an IBM study of CEOs. Apollo Global Management chief economist Torsten Slok found profit margins outside the tech sector show no sign of rising from AI yet, and industries built on heavy physical assets or regulation, healthcare, banking, manufacturing, and more, will need years of process rework before returns show up. (Learn more)
Boards are pushing anyway. Cisco surveyed 2,500 CEOs across 23 countries and found 69% see AI adoption as essential to staying competitive, yet more than half say their infrastructure could limit AI initiatives, and almost two-thirds worry they’re underinvesting. Infrastructure modernization ranks as the top 2026 priority for 40% of CEOs, and 72% still expect humans to keep oversight of AI systems through 2030. (Learn more)
Why it matters: Clients chasing AI returns on a timeline their own economist says doesn’t exist need a workforce partner who can show a number now, not in three years.
The true hiring story is the AI-and-defense buildout, not AI displacement
Global AI infrastructure investment hit $318 billion in 2025, up 81% in a single year, and Nvidia CEO Jensen Huang expects $3 to $4 trillion more in spend over the next four years. McKinsey puts the 15-year number at $6.7 trillion. Layer on global defense spending, which hit $2.9 trillion in 2025, with the US alone budgeted for $954 billion in FY2026 and $1.5 trillion in FY2027, and TAtech’s Peter Weddle calls it the “Big Bang in Talent Acquisition.” (Learn more)
Recruiters are already repositioning around it. Randstad CEO Sander van ‘t Noordende says vacancy rates for AI-adjacent roles like AI trainer and automation specialist top 25%, and demand for electricians, plumbers, and builders tied to data center construction is growing three times faster than professional roles. ManpowerGroup data shows AI-related US job postings up 95% year over year in the first half of 2026 while overall postings fell 16%. Robert Walters research says one in two AI roles in the UK and US could go unfilled by 2028, an 800,000-plus shortfall in the US alone. (Learn more)
Why it matters: This is the vertical to lean into this quarter. The general professional-hire market is still soft. Trades and technical desks tied to AI infrastructure and defense builds are where the demand is real and durable.
You outsourced the AI. The liability didn’t leave with it.
Companies deploying third-party AI tools are the ones courts and regulators go after, not the vendor that built the model. Harvard Business Review points to recent suits against Peloton, iTutorGroup, Workday, and Cigna as proof that accountability falls on whoever sits closest to the end user. The authors name four exposures most firms haven’t managed: opacity in the models they didn’t build, liability triggered by their own customization, dependence on vendors that are hard to replace, and regulations that differ by state. The fix starts with hard-wiring transparency into vendor contracts and anchoring compliance in a recognized framework like NIST’s AI Risk Management Framework or ISO/IEC 42001. (Learn more)
Why it matters: If an AI tool touches your sourcing or screening anywhere in the pipeline, the liability for what it gets wrong sits with you, not the vendor, and that’s worth an audit before it’s tested in court.
Your fastest fill and your top biller are hiding your real numbers
Speed still wins the order, but it stopped being a differentiator. Bullhorn’s 2026 GRID report found faster placement correlates with revenue growth more than almost any other metric, and top performers now place 56% of candidates in under 10 days. But ClearlyRated’s benchmark shows client satisfaction gains came almost entirely from timing: 92% of clients say firms deliver on schedule, while only 57% say the work is consistently high-quality and error-free. In StaffingHub’s own 2025 State of Staffing data, 31% of agencies now name quality of hire their top metric, ahead of cost-per-hire at 19% and time-to-fill at 18%. (Learn more)
Billings hide the same blind spot at the recruiter level. Loxo founder and CEO Matt Chambers argues that most firms can name their top biller but not their most profitable placement once fallout, data decay, and client overhead are counted. The metrics that actually predict business health are stage-to-stage conversion, gross profit per recruiter, and account-level free cash flow, not call volume or who closed the most deals. (Learn more)
Why it matters: Two unrelated pieces land on the same blind spot. The number your firm brags about, fastest fill or biggest biller, isn’t the number that predicts whether the account renews or the placement actually made money.
The StaffingHub Brief provides weekly insights for staffing agency leaders and publishes every Friday.
