While the implementation of artificial intelligence in the professional world is progressing, it has not yet led to a productivity revolution, according to a study conducted by a team of HEC professors and researchers. The tests show that the distinction between AI-generated texts and those written by humans often remains blurred, impacting employees’ perceptions and recognition of their work. Furthermore, the phenomenon of « hidden adoption » hinders beneficial use across an organization, raising challenges in terms of transparency, security, and intellectual property. Despite the enormous potential of Artificial Intelligence (AI) to transform the professional landscape, its adoption remains marked by paradoxes and shortcomings. Recent studies, notably from HEC, indicate that, although AI is gradually being integrated into businesses, it has not yet demonstrated a significant increase in productivity. Several obstacles, such as the recognition of AI-generated texts, biased perceptions among employees and managers, and unclear company policies, are hindering widespread adoption.Biased Perceptions of AI A study conducted by HEC revealed that many managers struggle to differentiate between an AI-generated document and a human-generated document. By testing texts produced with and without the help of ChatGPT, it emerged that 73% of managers did not notice the use of artificial intelligence, and 77% believed they saw it where it was not present. This confusion illustrates the difficulty of recognizing the influence of AI on content, impacting the perception of the added value of AI tools. Hidden Adoption at the Detriment of ProductivityA phenomenon known as « shadow adoption » illustrates the behavior of employees using AI-generatedAI toolswithout informing their superiors. This secrecy represents a major obstacle to the effective integration of AI into the production process. Although individuals can benefit from personal time savings, the lack of a transparent policy prevents more widespread use of AI, which could, in theory, improve the company’s overall productivity. The dangers of a lack of guidelinesThe HEC study suggests that without a clear transparency policy, companies are putting themselves at risk. The risks are multiple: from work quality to data protection, including intellectual property in creative sectors. The indiscriminate use of AI without rules could also foster misinformation and compromise security. Hence the importance of defining rules for the use of these technologies. Seniors, unexpected allies of AI Contrary to popular belief, it is often senior employees who are most inclined to adopt artificial intelligence. Their experience not only allows them to identify errors but also to better assess the risks associated with the use of AI tools. Rather than seeing AI as a threat to their jobs, these employees perceive the dangers related to the reliability of the work produced. A need to reassess incentivesTo overcome these obstacles, the recommendation is to readjust incentive mechanisms within companies. The idea is to recognize the efforts of employees using AI in a way that benefits everyone. For example, younger generations, particularly in the legal and architectural professions, are looking for employers that make optimal use of AI tools. It is therefore crucial to develop an open and constructive culture around these technologies to attract and retain talent and boost productivity.
