Recruiters and managers – almost all of them use AI, but there is a lack of guidelines

Writing, data analysis, information research and synthesis, and the automation of repetitive tasks. Whether it’s ChatGPT, Claude or Gemini, intelligent agents have now become part of the daily routine for the vast majority of managers and recruiters when it comes to these activities. The same applies to employees, who, however, are often not open about their use of these tools and find it difficult to discuss them with colleagues. What is lacking, in fact, are policies and guidelines governing their use and limitations, particularly in smaller organisations.

Nine out of ten respondents (87%) who took part in a Censuswide survey for Indeed – based on a sample of 1,000 people comprising recruiters, managers, employees and jobseekers in Italia – say they use generative AI. For a third (27 per cent), its use has become part of their daily routine. How? Almost always without clear guidelines, and this is increasingly the case, particularly as we move from large companies to small and micro-enterprises. The rapid adoption of AI, however, is presenting organisations with a new structural challenge. As Gianluca Bonacchi, Talent Strategy Advisor at Indeed, explains, the figures show that ‘Artificial Intelligence is no longer a future prospect, but a reality already present in Italian workplaces. The data shows that the speed at which workers begin using these tools often outstrips companies’ ability to establish guidelines. Creating a framework is essential to ensure individual efficiency and reduce risks for companies’.

The Censuswide survey for Indeed reveals that nearly two in three companies surveyed (65 per cent) already have a framework or guidelines in place regarding the use of artificial intelligence by employees. A quarter (23 per cent) have formal written regulations, whilst 42 per cent have opted for more informal internal recommendations. Just over a third (35%) of companies that currently lack policies and guidelines are planning to introduce them, confirming that the use of this technology is driving a review of business workflows. “The data highlights that employees still need guidance on the use of Generative Artificial Intelligence,” adds Gianluca Bonacchi, Talent Strategy Advisor at Indeed. “Defining how to use the tools, what data can be processed and how to document their use enables individual experimentation and usage to be transformed into a team competence, creating the environment needed to successfully drive digital transformation within the organisation.”

However, the pace of policy implementation varies considerably depending on company size: medium-to-large companies, with over 250 employees, are the best organised; in this case, 68 per cent have already adopted a set of rules. Among micro-enterprises with fewer than 20 employees, the figure is 45 per cent. The activities for which AI is most commonly used are content writing and optimisation (37 per cent), data analysis (36 per cent), information research and summarisation (35 per cent) and the automation of repetitive tasks (29 per cent).

Introducing clear company guidelines or policies helps to guide employees on how to use AI and also helps to prevent a lack of transparency. Only 32 per cent of managers and recruiters say that most employees feel comfortable openly sharing how they use AI in their work. Forty-seven per cent believe that only a few would do so, whilst 16 per cent think that many prefer not to disclose it. Among employees, nearly two in five (37 per cent) say they have used AI tools at work without informing their colleagues or manager. Furthermore, only 8 per cent of workers say they always make their use of AI explicit in professional contexts, whilst 22 per cent say they rarely or never do so. Workers experience uncertainty, fuelled by operational concerns: among the main concerns raised by survey participants are the fear of becoming too dependent on technology (25 per cent), issues relating to data privacy and confidentiality (24 per cent) and the fear of a reduction in human interaction (17 per cent).