Abstract:
As companies continue to increase their investment in generative artificial intelligence and continue to emphasize the efficiency improvements brought by AI to their employees, more and more professionals are beginning to face pressure to use AI tools. However, a new survey shows that many employees do not actually know how to use these tools effectively, but still deliberately exaggerate or even disguise their level of AI usage to cater to management expectations.
A study co-conducted by business intelligence company Visier surveyed 1,000 full-time U.S. employees. The results showed that nearly half of the respondents misreported their frequency and ability to use generative AI at work. Visier calls this phenomenon "Performative AI."
Data shows that 48% of employees surveyed admitted to exaggerating their experience or proficiency in using AI tools to colleagues or leaders. Interviewees said that the main reason behind this behavior is not to truly master the relevant technology, but to show their competitiveness in the AI era. At the same time, they generally believe that companies do not provide enough training and clear guidance, resulting in many people lacking confidence in actual operations even if they are required to use AI. About 45% of respondents said they feel pressure to use AI, but are not confident that they can use these tools effectively.
The survey also found that employees have obvious doubts about the company's AI strategy and management's cognitive level. Many believe that company executives do not truly understand the actual capabilities of AI and how it will impact workflows and job responsibilities.
Only 28% of respondents believe that employers fully understand how AI is used in work scenarios; only 29% believe that management truly understands employees’ concerns about AI; and only 33% believe that bosses understand the level of training required by employees. In addition, 32% of employees said that the company has not yet formulated a clear plan to evaluate the actual impact of AI on the organization, and another 27% even suspected that management has no relevant plan at all.
Rather than being forced to increase the use of AI, employees prefer to receive concrete support. The survey shows that about 32% of the respondents want more training opportunities and skills improvement resources; 31% want companies to be more transparent about AI strategies; 18% want management to more clearly explain how AI will affect their future work and career development.
In recent years, many companies have been influenced by technology companies such as Nvidia, Microsoft and OpenAI. They believe that AI can bring significant productivity improvements, and therefore have begun to actively monitor employees' AI usage. Last year, Microsoft launched related tools to help companies track Copilot usage data and designed it into a ranking and competition mechanism.
However, this approach does not always achieve the desired results. Earlier this year, there were cases within Amazon where employees artificially increased the use of AI tools in order to improve ranking results. Although the company has publicly stated that it will not evaluate employee performance based on AI usage, sources said that relevant usage data may still affect performance appraisals. On the other hand, language learning platform Duolingo has stopped evaluating employees based on their frequency of AI use this year.
Employees have mixed opinions on the impact of AI entering work processes. Some people who participated in Visier's survey said that AI promotes team collaboration and communication. But at the same time, many programmers have publicly complained that companies’ forced promotion of AI programming tools is weakening their professional skills. Some developers pointed out that management often does not understand the real work conditions of grassroots employees, and more and more of their time is spent checking and correcting low-quality code generated by AI instead of truly improving development efficiency.
The survey results reflect that there is still a clear gap between the current enthusiasm for AI in the workplace and its practical applications. Employees are not outright resistant to AI, but are forced to appear more proficient at AI than they actually are without training, guidance, and a clear strategy. This "performative AI" phenomenon is becoming a new issue that cannot be ignored when enterprises promote artificial intelligence transformation.
Comments