Gallup has released its State of the Global Workplace 2026 report, titled The Human Side of the AI Revolution. For anyone working in product, engineering, leadership, or organizational transformation, it is a useful reality check.
The main message is simple: AI tools may already be powerful, but productivity gains do not automatically appear just because a company adopts AI. They depend on how well people are led through change.
1. Employee engagement is falling again
According to Gallup, global employee engagement declined for the second year in a row in 2025, reaching 20% — the lowest level since 2020.
In other words, only one in five employees worldwide is truly engaged at work. The rest are either not engaged or actively disengaged.
Gallup estimates that low engagement cost the global economy around $10 trillion in lost productivity last year, equal to about 9% of global GDP.
For companies, this is an important signal. Productivity problems are not always caused by missing tools, weak processes, or lack of headcount. Often, the deeper issue is that people do not feel connected to their work, their team, or their organization.
2. AI improves individual productivity, but not always organizational performance
One of the most interesting findings in the report is about AI.
Among U.S. workers in organizations that have implemented AI, 65% say AI has had a positive impact on their personal productivity. However, only 12% strongly agree that AI has transformed how work gets done in their organization.
That gap matters.
AI can help individuals write faster, analyze faster, summarize faster, and automate routine tasks. But there is a big difference between “this tool helps me personally” and “our organization is now more effective.”
Gallup’s point is that this gap is not mainly about model quality. It is about management, adoption, and organizational readiness.
3. Managers are becoming the weak link in AI transformation
The report shows that the recent decline in engagement is driven largely by managers.
Since 2022, manager engagement has dropped by nine percentage points. In 2025 alone, it fell from 27% to 22%.
This is a serious problem because managers are the people who turn strategy into everyday behavior. If a company rolls out AI tools, but managers do not help teams understand how those tools fit into real work, AI remains “another tool” rather than a meaningful change in how the organization operates.
Gallup also found that employees whose managers actively support their team’s use of AI are dramatically more likely to believe AI has transformed how work gets done.
The takeaway: AI adoption is not just an IT initiative. It is a management challenge.
4. The job market looks slightly better, but AI anxiety is growing
Globally, employee perceptions of the job market improved slightly in 2025: 52% of employees said it was a good time to find a job.
At the same time, concerns about AI-related job losses are increasing. In the U.S., 18% of employees said it was likely that their job could be eliminated within the next five years because of technological innovations such as automation or AI. In organizations where AI has already been implemented, that number rises to 23%.
In some industries, the concern is even higher. Finance, insurance, and technology are among the sectors where employees report the strongest expectations of AI-related job disruption.
Interestingly, AI’s effect on employment does not appear to be uniformly negative. In large organizations, employees are more likely to report workforce reductions after AI implementation. In smaller organizations, employees are more likely to report workforce expansion.
AI is not simply “replacing people.” It is reshaping organizational structures.
5. Wellbeing has slightly improved, but stress remains high
Gallup reports that global employee wellbeing improved in 2025 for the first time in three years. The share of employees classified as “thriving” increased from 33% to 34%.
However, negative daily emotions remain above pre-pandemic levels. In the global summary:
- 40% of employees reported experiencing a lot of stress the previous day;
- 22% reported anger;
- 23% reported sadness;
- 22% reported loneliness.
This matters for technology teams. A team can look productive on paper while being emotionally overloaded. In an environment of constant change — AI adoption, restructuring, layoffs, new expectations — emotional resilience becomes part of operational performance.
6. What this means for engineering and product teams
For me, the strongest takeaway from the Gallup report is this: AI does not make management less important. It makes management more important.
You can buy the best tools, deploy GitHub Copilot, ChatGPT Enterprise, internal AI agents, and automated workflows. But if the team lacks clarity, trust, feedback loops, and shared rules for using AI, the impact will remain local and inconsistent.
AI accelerates people who already understand what they are doing. It does not automatically fix unclear priorities, weak leadership, or burned-out teams.
Practical takeaways
Here are a few things organizations can do now:
Do not roll out AI as “just another tool”
Explain which workflows are changing, why they are changing, and what success looks like.Train managers, not only individual contributors
If managers do not understand how AI supports the team, the team is unlikely to use it systematically.Measure impact, not only adoption
The number of AI tool users is not a business outcome. Look at cycle time, decision quality, delivery speed, customer impact, and team load.Talk openly about job concerns
If employees fear that AI is mainly a threat, they will not see it as a tool for growth.Treat engagement as a readiness metric
Engaged teams handle change better. Disengaged teams resist it — sometimes silently.
Final thought
Gallup’s report is a useful reminder that the AI revolution is not only about models, prompts, and automation. It is also about people, managers, and an organization’s ability to change.
The technology is already powerful enough to matter. The bottleneck is increasingly the human system around it.
Maybe the next big productivity breakthrough will not come from a new model release. Maybe it will come from companies finally learning how to lead the people who use these models.