Scientists have identified a "burnout loop" due to the introduction of AI

Artificial intelligence (AI) is increasingly being perceived in many companies as a cost-saving tool. Instead of expanding their staff, employers redistribute tasks among existing employees, expecting that the neural network will help them "increase efficiency." However, research shows that in practice the workload does not decrease, and in some cases even increases. This was reported by RBC on February 25.

Aruna Ranganathan and Sinzi Maggie Ye, researchers from the University of California, Berkeley School of Business, analyzed the work of a technology company with 200 employees over the course of eight months. The results, published in the Harvard Business Review, showed that the use of AI did not reduce the volume of tasks, but made the work more intense. Employees, relying on the help of technology, took on more responsibilities than they could fulfill, and eventually began to work longer, including outside of working hours.

One of the study participants admitted that he expected a reduction in workload, but in reality he began to work as much or even more. At the same time, the company's management did not oblige to use AI — the initiative came from the employees themselves. As a result, employees began to perform some of the tasks that had previously been outsourced or required the hiring of additional personnel on their own using neural networks.

An additional problem was the blurring of the line between work and leisure. Employees used AI during lunch breaks and after the end of the working day, which made recovery difficult. The researchers note that speeding up the execution of individual operations increased expectations, both on the part of management and on the part of the employees themselves, which is why the total volume of tasks continued to grow.

Programmer Siddhant Khare also spoke about the "productivity paradox" in an interview with Business Insider. According to him, AI reduces the cost of individual operations, but increases the cost of human labor for coordination, verification and decision-making. He noted that he wrote more code in a quarter than ever before, but felt more drained. Constantly switching between multiple tasks, each of which "takes only an hour with AI," creates a significant cognitive load.

Love is off the schedule: how work has become the main enemy of youth's personal life

Young people are increasingly sacrificing their personal lives for the sake of career and stability

Other researchers draw similar conclusions. According to the Massachusetts Institute of Technology, in 95% of companies, the introduction of generative AI has not led to a noticeable increase in productivity or revenue. One of the reasons is the difficulty of adapting models to specific business processes. At the same time, the AI Security Center (CAIS) reported that modern models are able to perform only a small part of the tasks of a remote employee without significant human refinement.

In Russia, the use of AI is also expanding. As Natalia Golovanova, head of the Superjob research center, noted, 10% of office employees use such technologies all the time, and about a third use them periodically. Neural networks are most often used for document management and data analysis.

"There are no mass layoffs due to the introduction of AI yet, but it is already affecting hiring: there are fewer vacancies for simple office tasks and positions for beginners," the expert explained.

Experts also point to other risks: a decrease in the quality of work, dependence on algorithms, a decrease in the number of internships and programs for beginners. According to Grigory Finkelstein, senior partner and head of HR Consulting at Ecopsi, AI creates the illusion of simplifying tasks, but without well-defined algorithms and controls, the result may worsen. Kontakt InterSearch partner Anastasia Ovcharenko adds that problems arise when employees try to compensate for the lack of competencies using a neural network.

According to analysts, in the coming years, companies will move from mass use of open tools to more highly specialized solutions embedded in corporate systems. At the same time, the labor market will increase the demand for specialists who not only have specialized expertise, but also know how to work competently with AI — formulate tasks, verify results and not shift responsibility for the final quality onto technology.

On January 20, it was reported that the demand for specialists with experience in the use of artificial intelligence (AI) in Russia has almost doubled — in 2025, the number of vacancies with such requirements has increased by 90%. According to the Avito Jobs study, AI skills were most often in demand among sales professionals.

All important news is on the Izvestia channel in the MAX messenger.