Businesses worldwide have poured investment into artificial intelligence, yet fewer than 40% report meaningful productivity gains. A new report from Top Employers Institute reveals why: nearly half of workers say they do not understand how AI is supposed to help them perform better, even as 96% of senior leaders expect it to do exactly that. The gap between expectation and reality is not closing, and as AI enters a far more powerful phase, the stakes of getting the fundamentals wrong are rising sharply.
Designing Work for the Age of Agentic AI examines what it will take for businesses to close that gap. Drawing on data from almost 2,500 certified organisations across more than 120 countries, the report finds that businesses which simply layer new AI capabilities onto old structures will see fragmented gains at best, and systemic disruption at worst.
A Category Moving Fast, With Risks to Match
By April 2026, approximately half of European organisations surveyed by Top Employers Institute were already using agentic AI in some form, a trajectory that outpaces the governance and workforce structures most businesses currently have in place. While this is a signal for the rest of the world, for many workers, the reality on the ground is concerning. AI deployment has been linked to rising burnout, loneliness and growing anxiety around job security and surveillance. There is also evidence that employees are beginning to use AI agents in unauthorised ways to bypass organisational guardrails in pursuit of speed, creating new and largely unmonitored risks.
Sandra Botha, Lead HR Auditor at Top Employers Institute, says: “Technology is moving faster than most organisations’ ability to absorb it. The next challenge is not deploying more AI. It is designing coherent systems in which humans and technology operate together productively and safely, while managing the very real risks to employee wellbeing, morale and organisational control.”
Three Principles for Getting It Right
The report organises its guidance around three principles:
- Build clarity and confidence. When AI is deployed without clear communication or visible benefit to employees, the result is not outright resistance but something harder to diagnose: limited adoption, quiet workarounds and a gradual erosion of confidence in AI-supported decisions. Productivity gains are increasingly perceived as organisational wins rather than shared value. Top Employers that have seen rising employee engagement are those that deliberately apply AI in ways that support their people’s careers, involve employees early in the integration process, and continuously evaluate the impact on both business outcomes and employee experience. When employees feel AI is working for them, not just around them, adoption deepens, and the efficiency and morale gains that AI promises begin to materialise. These organisations are 5% more likely to be using AI for strategic workforce planning and 5% more likely to be using it for operational efficiency.
- Rethink roles, rewards and workforce planning. Agentic AI is giving rise to what the report calls the “AI Orchestrator”: a generalist worker who uses AI to operate with the reach and output of a small team, coordinating workflows, reducing handoffs and translating intent into execution. This is reshaping how employees experience their day-to-day work, often for the better, but only when organisations have deliberately redesigned roles and structures around the shift. Yet most job architectures, performance frameworks and pay structures were built for a different model. Without deliberate redesign, businesses risk misaligning reward with contribution and undermining the morale gains that well-deployed AI can deliver. The skills and planning gap remains considerable: only 61% of Top Employers have a formalised process for identifying long-term workforce and skills needs, and only 41% consider multiple future scenarios. Top Employers with rising engagement scores are 8% more likely to have strong strategic workforce planning in place and 9% more likely to use scenario planning.
- Harness ethical foundations for competitive advantage. As AI agents coordinate across workflows and interact with other systems, risk shifts from isolated errors to systemic vulnerabilities, making governance not just a safeguard but a strategic enabler. Among Top Employers, ethical AI frameworks have moved from 42% to 62% adoption in just one year, while responsible AI deployment practices have risen from 48% to 68% over the same period. Top Employers that list innovation as their number one business priority are more likely than peers to have already established an ethical AI framework, suggesting that the organisations moving fastest on AI are also the ones investing most deliberately in governance. Organisations that have already established clear principles and boundaries are scaling agentic systems with greater confidence, because the rules of engagement and the credibility they create are already in place.
“The next chapter of work will be shaped by how organisations design and govern intelligent systems. Roles will stretch, workflows will shift, and long-held assumptions about accountability, contribution and trust will come under pressure,” notes Botha. “Organisations that will lead in this next phase are not necessarily those with the most advanced technology. They are the ones that have thought hardest about how people and systems work together.”
The organisations gaining the greatest value from AI are not simply deploying new tools – they are redesigning workflows, roles and workforce strategy around them.
Download Designing Work for the Age of Agentic AI fromhttps://www.top-employers.com/blog/designing-work-for-the-age-of-agentic-ai/ toexplore the practical frameworks, workforce insights and governance principles shaping the next generation of high-performing organisations.
