AI adoption in the American workplace has crossed a critical threshold, but the productivity gains are landing unevenly. New survey data reveals a widening gap between managers and frontline employees, raising concerns that AI tools may be entrenching a two-tier workforce rather than lifting overall performance.
The data that got my attention
For the first time, half of U.S. workers say they use AI at work at least a few times a year. But beneath that milestone lies a fracture: 21% of leaders report AI has had an “extremely positive” impact on their productivity, compared to just 13% of individual contributors. That gap — roughly 60% — is the clearest signal yet that AI is creating a two-tier workforce within organizations, not across them.
Gallup’s Q1 2026 survey of 23,717 U.S. employees found that AI adoption has crossed the halfway mark, with 13% using AI daily and 28% using it a few times a week. Yet the benefits are not spreading evenly. The people at the top of the hierarchy are seeing the biggest gains. The people at the bottom are seeing the least.
Why this matters now
The AI productivity divide mirrors an existing fault line: knowledge work versus service work. Leaders and professionals in technical roles can plug AI into analysis, communication, and planning — tasks where AI excels. Workers in service and administrative roles have fewer clear use cases. When 89% of corporate leaders report no organizational productivity impact from AI (per NBER research of U.S., U.K., Germany, and Australia executives), even as individual leaders report personal gains, the picture becomes clear: AI is helping the people who direct work, not necessarily the people doing it.
This matters because the same organizations investing billions in AI are also reporting record-low employee engagement. Global engagement fell to 20% in 2025, the lowest since 2020. If AI benefits concentrate at the top while the rest of the workforce feels unchanged or threatened, engagement will fall further.
What the research actually shows
The Gallup data reveals several patterns that define the AI two-tier workforce:
| Metric | Leaders | Individual Contributors |
|---|---|---|
| Report “extremely positive” AI productivity impact | 21% | 13% |
| Work in remote-capable, knowledge-based roles | Most | Fewer |
| Clear AI use cases (analysis, planning, communication) | High | Mixed |
| Job elimination fear (overall U.S. average) | — | 18% (23% in AI orgs) |
Among employees who use AI, healthcare workers and technical professionals report the strongest productivity gains. Service workers and administrative support workers are more likely to say AI has had little, no, or even a negative effect on their productivity.
Large organizations tell an even starker story. In companies with 10,000 or more employees that have adopted AI, 33% are reducing their workforce compared to 30% expanding. In similar-sized organizations without AI, the pattern flips: 36% are hiring and 23% are cutting. AI adoption in large employers is now correlated with workforce contraction.
Only about 10% of employees in AI-adopting organizations strongly agree that AI has transformed how work gets done. The benefits are real but concentrated at the individual task level — drafting, summarizing, generating ideas — not at the organizational level.
A practical framework for leaders
Closing the AI productivity divide requires intentional action, not just broader tool deployment:
- Audit who benefits. Map AI productivity gains by role, department, and seniority. If the gains cluster at the top, your AI strategy has a distribution problem, not an adoption problem.
- Build role-specific use cases. Service and administrative roles need different AI applications than knowledge workers. Stop measuring success by whether people use AI and start measuring whether it helps them do their jobs.
- Address job security directly. Twenty-three percent of employees in AI-adopting organizations fear their job will be eliminated within five years. Unspoken fear kills engagement and adoption. Name it, address it, and connect AI to reskilling pathways.
- Invest in manager-led adoption. Employees whose managers actively support AI use are 8.7 times more likely to say AI has transformed how work gets done. Yet less than a third of employees say their manager supports AI use.
- Redesign workflows, not just tools. The gap between individual and organizational productivity gains exists because most organizations have not redesigned processes around AI. They bolted AI onto existing work and hoped for transformation.
The bottom line
AI adoption is wide but benefits are narrow. Half the workforce uses AI, but only 10% say it has transformed how their organization works. Leaders gain more than the people they manage. Knowledge workers gain more than service workers. Large employers with AI are more likely to cut jobs than expand. This is not a technology problem. It is a leadership and design problem. Organizations that distribute AI benefits across all roles — not just the ones with corner offices — will pull ahead. Those that do not will widen an engagement gap that is already at record lows.
Where to go from here
The AI productivity divide is a leadership challenge before it is a technology challenge. Leaders who assess their organization’s AI readiness across every role, not just the ones already benefiting, will build a workforce that adapts instead of fractures. If your team is navigating this transition, the AI Leadership Readiness Assessment can help identify where your adoption gaps are widest and what to fix first.
