“Why teaching employees to use AI may be more effective than replacing them with AI: A review of current evidence.”

By Steven Campos

AI will replace humans; is the first thing everyone says when it comes to AI. Other things you often hear is “Why pay a human to do the job when you can get an AI to do the same job for free?”. Recently ChatGPT specifically has seen a boom in recent years ever since it went public in November 2022. Now the conversation has turned to the topic of “It’s only a matter of time where AI will be taking our jobs”. But do companies really benefit from replacing them with AI? Everyone fears the future where companies do not want to hire humans in favor of using AI, however there isn’t enough empirical evidence from research that shows that companies that replace AI perform better than companies that use human employees. However, there is empirical evidence to suggest that companies might benefit from training employees to use AI in their job rather. Which begs the question, “If a company benefits from training employees to use AI, would that be more effective than replacing them with AI”

      To understand my argument, think of it like this it’s like machines, everyone was worried that machines would replace humans and take our jobs. However, machines have been beneficial to use. It has created a whole new jobs called “IT”, “Engineers”, “Research and development” and all of a sudden people are not worried about robots taking our jobs but instead now see a whole new world with brand new jobs that people can now potentially have, and the same could one day go for AI and the direction it might one day go in the future.

Let’s first address the elephant in the room. Skills needed to learn to effectively use AI. If a company wanted to teach their employees to use AI they would need to teach them Human AI skills which is best explained as “Human AI skills refer to the ability and skills of people to interact and collaborate with AI technology (Mikalef & Gupta, 2021).” Within Human AI there are Technical AI skills, and Managerial AI skills. In the research paper by Meng An, Jiabao Lin, and Xin Luo titled “The impact of human AI skills on organizational innovation: The moderating role of digital organizational culture” they came to this conclusion after their findings “First, we found that human AI skills positively affect organizational innovation. Specifically, both technical AI skills and managerial AI skills have positive impacts on organizational innovation.” This conclusion shows that companies might benefit from teaching their employees on using AI effectively.

Just merely using AI is not the same as understanding on how to use AI. This is something that in the research paper titled “Integrating AI in organizations for value creation through Human-AI teaming: A dynamic-capabilities approach “ by Cristina Simon, Elena Revilla, and Maria Jesus Saenz they mentioned in their conclusions “to operate successfully, AI requires much more than “final users.” It requires professional wisdom from various angles to learn and correct itself through continuous improvement. Thus, people engagement in a project is key, and managers should create a climate of transparency in which team members exchange their experiences and build trust in AI and its potential benefits. Along these lines, managers will have to leverage team members who are more prone to AI, as well as make visible the value that the department will derive from the AI outcome throughout the project and communicate how it will compensate for the effort invested.” For full context, the research paper focused on a Santander bank in Spain where participants for the sake of the study used AI. In the study that they conducted, the team went back and forth with the AI by giving it sensitive information about the company as if it was an employee albeit as they described in their own words “an intern that saves us of the tedious tasks”. Even though in this research paper, the AI felt like to the team they could just simply replace interns with AI. What if we were to take to the next level, what if we replaced an actual employee with an AI, would the company or group still benefit like the team in this research paper?

In the research paper titled “Let me transfer you to our AI-based manager: Impact of manager-level job titled assigned to AI-based agents on marketing outcomes” by Yongwoong Andrew Jeon, in his conclusion had this to say ““AI service employees” should be deployed in a different way—that is, the positive perception of the AI manager was contingent on whom the AI manager was seen to be paired with (the human representative versus the AI representative)”, even though his research was mainly about replacing customer service employee with AI, he noted that the AI did perform better when paired with a human. I mention this, because it is important to note that AI still needs humans which is what I am making the case for why companies should teach their employees how to use AI. However, what if we took this to the next level. By having AI fully replace a customer service employee? Luckily in one research paper titled “Can AI really help? The double-edged sword effect of AI assistant on employees’ innovation behavior” by Meng Yin, Shiyao Jiang, and Xiongying Niu, they wanted to see if by using AI does that negatively or positively affect the employees innovation ability. Here were their results “Based on TMS, we investigated the double-edged sword effect of AI assistants with high intelligence on employees' innovation behavior. Using two experimental scenarios, the results indicate that (1) an AI assistant characterized as high-intelligence (vs. low-intelligence) has a positive indirect effect on employees' AI-enabled innovation behavior via creative self-efficacy, and the indirect effect is stronger when organizational AI readiness is higher than it is lower; and (2) an AI assistant, characterized by high-intelligence (vs low-intelligence), has a negative indirect effect on employees’ AI-enabled innovation behavior via STARA awareness when organizational AI readiness is lower.” Not only in their research did they find that AI did have a positive affect on employees, they also found that depending on the situation AI also has a negative affect on employees. Meaning that just simply using AI doesn’t always guarantee positive results, and it is like they mentioned a “Double edged sword” when it comes to using AI. But after reading this research paper, one question I pondered was if AI is bad. What if we got rid of it? Luckily one country did ban ChatGPT, Italy in 2023.

In the research paper titled “The Impact of Generative AI on Information Processing: Evidence from the Ban of ChatGPT in Italy” by Jeremy Bertomeu, Yupeng Lin, Yibin Liu and Zhenghui Ni. They wanted to see the effects of what would banning ChatGPT have on the country Italy since they banned it. Here is what they found “These negative effects are more pronounced for analysts relying more on ChatGPT before the ban or with a background in technology. These findings collectively suggest that the ban reduces analysts’ information processing capacities and, therefore, their function as information intermediaries.” It is interesting to note that banning ChatGPT did not have positive effects on the country, but it had a negative effect. To me this tells me that people are underestimating AI, people are so focused on AI replacing people that research like this can show us that humans using AI can often be better. But is there a scenario where humans using AI can lead to the opposite and be worse?

Researchers in one study in the research paper titled “Preventing promotion-focused goals: The impact of regulatory focus on responsible AI” by Samuel N. Kirshner, and Jessica Lawson. They conducted a study where employees who were using AI had an incentive to be successful, the incentive was a promotion. This is what they concluded in their research “Our results across all studies consistently show that an organization adopting a promotion-focused orientation decreased participants' stated ethicalness of the target RAI values and increased their likelihood of engaging in AI-based UPB. Furthermore, the RAI values act as a novel mediator explaining the relationship between the firm's regulatory focus and the decision to implement the unethical AI.” That employees who were focused on being successful missed used AI in an unethical standard. This research highlights that while companies do benefit from humans using AI better than just replacing humans with AI. It can sometimes come at a cost, employees are sometimes under pressure to perform good in order to get a good evaluation or get that promotion and can lead them to missing using AI in unethical ways that benefit them by making them appear more successful than they actually are to appear as a more valuable employee to the company.

In conclusion, AI is a hot topic of discussion. And it is no surprise that everyone is worried about a future where AI replaces humans. Hopefully in my findings I was able to show that maybe replacing humans with AI might not be a good idea and why maybe teaching human employees to use AI might be more effective. While looking for these research papers and reading them I did find some gaps in the research. While yes, this research did in fact find that using AI was beneficial, the research paper that conducted research on employees miss using AI in unethical ways. Raise questions that I believe might need more research to be done in. Are people or employees who are using AI and are getting positive results. Are they getting positive results at the cost of engaging in unethical behavior with AI to get the positive results that they want?

Bibliography

Cristina Simón, Elena Revilla, Maria Jesús Sáenz, Integrating AI in organizations for value creation through Human-AI teaming: A dynamic-capabilities approach, Journal of Business Research, Volume 182, 2024. https://doi.org/10.1016/j.jbusres.2024.114783.

Meng An, Jiabao Lin, Xin (Robert) Luo, The impact of human AI skills on organizational innovation: The moderating role of digital organizational culture, Journal of Business Research, Volume 182, 2024, 114786, ISSN 0148-2963, https://doi.org/10.1016/j.jbusres.2024.114786.

Yongwoog Andrew Jeon, Let me transfer you to our AI-based manager: Impact of manager-level job titles assigned to AI-based agents on marketing outcomes, Journal of Business Research, Volume 145, 2022, Pages 892-904, ISSN 0148-2963, https://doi.org/10.1016/j.jbusres.2022.03.028.

Meng Yin, Shiyao Jiang, Xiongying Niu, Can AI really help? The double-edged sword effect of AI assistant on employees’ innovation behavior, Computers in Human Behavior, Volume 150, 2024, 107987, ISSN 0747-5632, https://doi.org/10.1016/j.chb.2023.107987.

Jeremy Bertomeu, Yupeng Lin, Yibin Liu, Zhenghui Ni, The Impact of Generative AI on Information Processing: Evidence from the Ban of ChatGPT in Italy, Journal of Accounting and Economics, 2025, 101782, ISSN 0165-4101, https://doi.org/10.1016/j.jacceco.2025.101782.

Samuel N. Kirshner, Jessica Lawson, Preventing promotion-focused goals: The impact of regulatory focus on responsible AI, Computers in Human Behavior: Artificial Humans, Volume 3, 2025, 100112, ISSN 2949-8821, https://doi.org/10.1016/j.chbah.2024.100112.