A new survey reported by TechCrunch indicates that more than half of managers say it has become harder to evaluate employees’ work after adopting AI tools. While the article points to a growing operational problem rather than a security or regulatory incident, it highlights an important issue for European businesses moving AI into daily workflows: productivity gains may be easier to promise than to measure.

What happened

According to the survey cited by TechCrunch, many managers say AI adoption is complicating how they assess employee performance. The reported finding is that more than 50% of managers are finding it harder to evaluate work after AI tools were introduced.

The source points to a common workplace reality around generative AI: employees may produce drafts, research summaries, code, marketing copy or internal documents faster, but it can become less clear how much of the final result reflects individual skill, how work quality should be judged, and which tasks still provide a meaningful measure of performance.

Why it matters for European businesses

For SMEs and mid-sized organisations, AI adoption is increasingly framed as a route to efficiency. But this survey suggests that deployment creates management and process challenges alongside productivity gains. If companies do not adapt their performance frameworks, they may struggle to distinguish between genuine improvements in output and changes caused mainly by AI assistance.

This matters in several practical areas:

• Work quality control: Faster output does not automatically mean better output. AI-generated material may still require review for accuracy, tone, compliance and brand fit.
• Accountability: Managers may need clearer rules on what employees are responsible for when AI is used in writing, coding, analysis or customer-facing content.
• Skills assessment: Traditional evaluation methods may become less useful if staff rely heavily on AI for first drafts or routine tasks.
• Training and governance: Businesses may need to define where AI is helpful, where human review is mandatory, and which tasks should not be delegated to AI tools.

For European employers, this is also relevant in regulated or client-facing environments where accuracy, documentation and human oversight matter. AI can improve speed, but poorly governed use can create quality risks, reputational problems or internal confusion about expectations.

Who may be affected

• Marketing teams using AI for copywriting, campaign planning, SEO support or content production.
• Software and IT teams using code assistants or AI tools for documentation and support tasks.
• Customer operations teams using AI for emails, summaries, knowledge retrieval or service workflows.
• Founders and SME leaders trying to link AI spending to measurable business outcomes.
• HR and department managers responsible for performance reviews, training and productivity management.

What companies should consider

Businesses introducing AI into everyday work may want to review not just the tools they use, but the management systems around them.

• Redefine performance metrics: Measure outcomes such as accuracy, conversion impact, customer satisfaction, delivery speed and error rates, rather than only volume of output.
• Document acceptable AI use: Create internal guidance on when AI can be used, what must be checked by humans, and how sensitive data should be handled.
• Separate drafting from decision-making: AI may assist with first drafts or routine tasks, but final approval and accountability should remain clear.
• Train managers as well as staff: AI adoption is not only a tooling issue. Managers need practical ways to evaluate work quality in AI-supported environments.
• Track business value directly: Instead of assuming AI improves productivity, monitor project completion times, quality issues, campaign performance and operational cost changes.

The broader lesson for European businesses is that AI implementation is not only a technology decision. It also changes how work is supervised, reviewed and rewarded. Companies that treat AI as a management and governance issue, not just a software purchase, are likely to be in a stronger position to capture real productivity gains.