Less than one-third of global workers feel their jobs are safe, ADP survey data shows

ADP's People at Work 2026 survey of more than 39,000 workers across 36 markets found only 22% globally strongly feel their jobs are safe, while 62% report up to five unpaid hours per week and 12% say they work 16 or more unpaid hours. Daily AI users were four times more likely than nonusers to say they were less productive, but they also showed higher engagement, lower stress and stronger team sentiment. Overall engagement remains low at 19% globally, with notable market gaps ranging from 29% in Brazil to 11% in China.

Analysis

The first-order read is not “AI demand is booming,” but that adoption is turning into a labor-arbitrage tool before it becomes a durable productivity story. If workers are using AI to compress tasks while still logging unpaid time, management may be capturing short-term cost leverage without seeing a clean boost in measured output, which is a warning sign for future burnout, attrition, and hidden re-hiring costs. That dynamic is favorable for software vendors monetizing usage, but less immediately positive for companies that expect AI to expand margins through headcount reduction alone.

For ADP specifically, the more important signal is that trust and engagement remain weak enough to sustain outsourcing demand. In a world where internal HR teams are asked to do more with less and employees feel less secure, payroll, compliance, scheduling, and benefits administration become more mission-critical, not less. The second-order effect is that the labor market’s softness may support sticky recurring revenue even if headline employment stays resilient, because firms will prioritize variable-cost, outsourced workforce management over fixed internal overhead.

The contrarian angle is that the market may be underestimating how slowly AI productivity actually shows up in financial statements. Daily usage can coexist with lower perceived productivity because the tool exposes complexity faster than it removes it; that means the “AI margin expansion” trade is probably too early in the cycle for most adopters, while the better near-term beneficiaries are infrastructure and workflow layers that sell picks-and-shovels. On the downside, if engagement keeps eroding, the eventual catalyst is not macro unemployment but a rise in turnover and management churn over the next 2-4 quarters, which would pressure firms with labor-intensive operating models first.

Key Decisions for Investors

• Long ADP on 6-12 month horizon: the setup favors sticky payroll/compliance spend as labor anxiety and HR complexity stay elevated; use pullbacks to build, with downside limited by recurring revenue quality and upside from defensive rotation in enterprise software.
• Pair trade: long ADP / short a labor-intensive services employer basket over 3-6 months, targeting margin divergence if engagement weakness translates into higher turnover and replacement costs; stop if wage growth reaccelerates sharply.
• Initiate a basket long in AI workflow beneficiaries with clear usage monetization (e.g., MSFT, NOW) versus short high-expectation AI adopters where productivity proof is weakest; 6-9 month horizon, as monetization should outrun cost takeout narratives.
• Avoid chasing broad AI-enablement beneficiaries that depend on immediate headcount reduction; instead consider selling upside calls against existing positions after earnings if management commentary leans on “efficiency” without quantified savings.
• Watch for any management commentary on retention or burnout in HR/service-heavy names; if turnover ticks up over the next 1-2 quarters, add to longs in automation/outsourcing platforms and fade employers with thin labor buffers.