AI Adoption Leaves Managers Scrambling to Measure Work, Survey Finds

A Nielsen survey of 1,700 directors and senior managers across seven countries, commissioned by Monday.com, found that more than half of respondents now find it harder to evaluate employee output after adopting AI tools. The report notes that 46% of managers use AI to inform decisions and 39% use it to draft performance reviews, while 58% expect to oversee more AI agents than human staff in the next year.

Key points

• Over 50% of managers struggle to assess employee output after AI adoption.
• 94% of managers are excited to use AI to improve work.
• 58% anticipate more AI agents than human employees next year.
• Nearly half spend extra time reviewing team output.
• 57% report increased tool proliferation.

What happened

The Nielsen survey, conducted in 2026 and covering seven countries, was commissioned by Monday.com to gauge the impact of AI on managerial workflows. Respondents reported that polished AI‑generated content now arrives directly in managers’ inboxes, obscuring the source of the work. This has made it more difficult for managers to determine how much of a deliverable originated from an employee versus an AI assistant. Consequently, almost half of the managers surveyed say they spend more time reviewing and guiding their teams’ output than before.

The study also found that 46% of managers use AI to help shape performance judgments, and 39% rely on AI to draft performance reviews. The report warns that large language models (LLMs) can introduce bias, echoing research from the Luxembourg Institute of Socio‑Economic Research and a 2025 paper that found AI models can be biased against black male candidates. In addition, 58% of managers expect to oversee more AI agents than human employees in the coming year, indicating a shift toward AI‑centric supervision.

The data also shows a surge in tool usage. Fifty‑seven percent of managers say their organizations now employ more tools than before, and about 51% report that these tools have become more interdependent and clustered. Only 29% describe their AI infrastructure as fully optimized and centralized.

The backstory

The enthusiasm for AI mirrors the early‑2000s internet boom, according to Michal Miller Levi, head of market research at Monday.com. Levi compares the current “magic” to that period, noting that people see AI as a powerful new asset. However, she points out that AI is evolving faster than organizational processes. “AI works very well for managers individually, but they also have a different role as managers. They need to drive the organization as well … Their struggle is due to the fact that AI is just moving faster than the organizational processes,” she told TechCrunch.

Businesses had expected AI to streamline software stacks, but the survey found the opposite: the number of tools in use has increased, and many employees are adopting unapproved AI applications. This trend has created a new layer of invisible labor for managers, who must decide which tasks to delegate, monitor AI outputs, and intervene when errors occur.

Analysis: Managerial overload and bias

The core problem is attribution. When AI‑generated content is polished and delivered as a finished product, managers lose visibility into the creative process. This forces them to spend more time dissecting work and verifying authenticity. The additional time required for review is a direct cost that can erode productivity gains promised by AI.

Bias compounds the issue. Because LLMs learn from existing data, they can reproduce human patterns and reinforce existing inequities. When managers employ AI to draft performance reviews, the risk of amplifying bias rises, especially if the AI is not trained on diverse datasets. The survey’s findings that 46% of managers use AI for decision‑making and 39% for reviews suggest that bias could become systemic if unchecked.

Analysis: Opportunity for governance and new metrics

The findings also reveal an opportunity. Managers are already shifting evaluation criteria: 78% of respondents said they now weigh prompt engineering, task delegation, and experimentation more heavily. These new metrics align with AI usage and could provide a framework for fair assessment. However, without standardized attribution mechanisms—such as metadata tagging or version control—managers will continue to struggle.

Centralizing AI tool inventories and enforcing governance can reduce tool proliferation. The survey shows that only 29% of organizations have fully optimized AI infrastructure. By creating a single source of truth for AI applications, companies can limit shadow usage, improve security, and streamline oversight.

What builders and businesses should do

• Implement transparent attribution for AI‑generated content, such as metadata tags or version histories.
• Treat prompt engineering as a formal skill and include it in performance metrics.
• Centralize AI tool inventory and enforce strict governance policies.
• Provide bias‑awareness training for managers and set up safeguards for automated reviews.
• Pilot AI‑augmented performance frameworks before scaling across the organization.

The bottom line

AI’s promise of productivity is tempered by managerial headaches. Companies that establish clear attribution, governance, and bias mitigation today will avoid costly misattribution and reputational risk. Watch for emerging regulatory guidance on AI‑assisted performance reviews in 2027, when compliance requirements are expected to tighten.