Discussions about artificial intelligence completely replacing programmers and triggering mass layoffs in the IT sector have recently become particularly loud. However, InfoWatch President and head of the association "Domestic Software" Natalya Kasperskaya has presented a reasoned position that questions this alarmist scenario.
According to her assessment, completely replacing humans with machines in software development is an economically inefficient strategy. Instead of professions disappearing, we are witnessing a transformation of roles. Programmers are increasingly less likely to write code manually and more often take on functions of managing AI: setting tasks, decomposing complex projects, and verifying results. Specializations, as Kasperskaya aptly put it, are "mutating" under the influence of technology, but not disappearing.
The key factor ensuring sustained demand for developers is the fundamental limitations of generative AI. Modern models are prone to hallucinations — producing confident but erroneous results. As long as machines can make mistakes, a belt of qualified specialists responsible for quality control and risk management will form around them.
Statistics from the US paint a different picture
Kasperskaya's optimistic forecast comes into obvious conflict with fresh data from the US labor market. According to calculations by the consulting firm Challenger, Gray & Christmas, in May 2026 alone, AI caused 38,579 layoffs in the US. This is a record monthly figure since observations began in 2023, accounting for 40% of all job cuts that month.
The numbers look alarming:
- May 2026: 38,579 layoffs due to AI.
- Full year 2025: 54,836 layoffs.
- Since the start of 2026: already 87,714 layoffs, exceeding last year's total.
The wave of layoffs has extended far beyond Big Tech: fintech companies and banks, including Standard Chartered and Coinbase, are announcing job cuts. Notably, not only rank-and-file employees are affected. Citadel hedge fund head Ken Griffin publicly acknowledged that agentic AI is already replacing highly qualified experts with master's and doctoral degrees, performing in hours work that previously took weeks.
Nevertheless, Kasperskaya's thesis finds support in the West as well. Some experts, including David George from Andreessen Horowitz and economist Tyler Cowen, call talk of the imminent disappearance of professions a myth. a16z research has not yet found evidence that AI leads to mass unemployment: changes are concentrated on individual tasks rather than covering entire professions.
My analysis: The paradox of the situation is that both sides are right. The layoff statistics are real and reflect the optimization of business processes. However, Kasperskaya correctly points to a long-term trend: AI is not destroying the programmer profession but creating new roles for "operators" and "architects" of neural networks. The labor market is undergoing a painful but inevitable restructuring, not a collapse.