AI's Unfulfilled Promise: Managers Grapple with Productivity Headaches

While artificial intelligence (AI) tools are being widely adopted by companies to boost employee productivity, a recent Nielsen survey, commissioned by work-management company Monday.com, reveals that these tools are simultaneously introducing significant new challenges for managers. More than half of senior leaders surveyed across seven countries reported finding it harder to evaluate employees' work after AI adoption. This difficulty stems from the fact that much of the output now reaches managers in a highly polished state, obscuring the distinction between human and AI contributions. Consequently, nearly half of managers are now spending more time reviewing and guiding their teams' output.

Despite these emerging difficulties, managers express strong enthusiasm for AI, with 94% excited to leverage it for improved work efficiency. Many are confident in identifying areas where AI can deliver maximum value, an excitement Michal Miller Levi, head of market research and customer insights at Monday.com, likened to the early days of the internet. However, Levi points out that the core issue is AI's rapid development outpacing companies' ability to adapt their organizational processes. Managers, while benefiting individually from AI, struggle in their role to drive organizational objectives when AI advancements move too quickly for existing structures.

The integration of AI has also prompted a re-evaluation of performance metrics. A significant 78% of respondents have altered their employee evaluation methods, now prioritizing skills such as prompt creation, effective delegation of tasks to AI, and experimental approaches to work. Yet, a fundamental question remains: how fairly can employees be judged when the precise origin of their work – human or AI – is ambiguous? Furthermore, this complexity is compounded by managers themselves increasingly relying on AI for their own decision-making; 46% use AI to inform decisions, and 39% employ it to draft performance reviews. This practice carries the risk of embedding new biases into the review process, as large language models (LLMs) are essentially prediction machines that can replicate human patterns found in their training data, as indicated by research showing potential biases against specific social groups.

The daily life of a manager is also being reshaped by AI, involving new responsibilities like deciding which tasks to assign to AI agents, monitoring their output, and intervening when issues arise. This oversight constitutes a new layer of