AI Enhances Productivity but Challenges Managers' Employee Evaluation and Workflow Control

Image: TechCrunch · Source

A new survey reveals AI tools boost productivity but complicate managers' ability to assess employee contributions and control AI-driven work processes, highlighting organizational adaptation challenges.

The adoption of artificial intelligence tools in workplaces is enhancing overall productivity, yet managers are encountering significant challenges in evaluating employee work and managing AI-driven outputs, according to a recent Nielsen survey commissioned by Monday.com. The survey polled 1,700 directors and senior managers across seven countries and found that over half of senior leaders are finding it harder to assess how much of employees' polished work derives from the employees themselves versus AI assistance.

This difficulty stems from AI tools often delivering refined results directly into managers' inboxes, blurring the line between human effort and AI-generated content. Consequently, nearly half of the surveyed managers report spending increased time reviewing and guiding their teams’ work compared to prior to AI integration. Despite these hurdles, 94% of managers expressed enthusiasm for leveraging AI to improve work and feel confident identifying high-value AI use cases within their roles. Michal Miller Levi, leading market research at Monday.com, compared this excitement to the early 2000s internet era, noting a similar perception of AI's "magic."

However, Levi emphasized that AI's rapid development outpaces organizational processes and adaptations. While AI works well for individual managers, their broader organizational responsibilities require new approaches. Reflecting this, 78% of managers have revised employee evaluation criteria to emphasize skills such as effective AI prompting, delegating tasks to AI, and innovation in workflow.

The challenge remains in fairly assessing employee performance when the AI's role is opaque, especially as 46% of managers use AI to assist decision-making and 39% employ AI-generated drafts for performance reviews. This practice may introduce bias, as large language models (LLMs) replicate patterns from their training data, sometimes perpetuating social prejudices. Studies cited in the survey point to LLM biases against certain social groups, underscoring the need for careful governance.

Additionally, AI adoption is altering managerial workloads. Supervising AI agents—including deciding task delegation, reviewing outputs, and intervening when issues arise—adds a layer of "invisible labor." Software complexity is also increasing; 57% of managers note more tools in circulation, with over 50% reporting an expanding software stack contrary to expectations of consolidation. Shadow AI use outside official systems further complicates control.

Looking ahead, 58% of managers anticipate overseeing more AI agents than human employees within a year, and 46% fear losing control over autonomous AI actions. There is concern about accountability, with 35% apprehensive about responsibility for AI-driven decisions. Managers generally prefer AI assistance with human involvement over fully autonomous agents.

Overall, the survey highlights that while AI presents productivity opportunities, its integration poses significant governance challenges. Without evolving management practices and organizational structures, managers may struggle to harness AI's benefits effectively.