Summary of Key Points
Many companies have resorted to layoffs under the guise of “AI transformation,” only to realize that AI cannot completely replace humans. AI requires significant human oversight and lacks certain capabilities, such as emotional intelligence and internal company knowledge. As a result, more than half of these companies have regretted their decisions and have rehired some of the laid-off employees. However, rehiring does not necessarily mean a victory for the employees; it often comes with changes in working conditions, reduced benefits, and increased workload. Meanwhile, Europe has established regulations and agreements to ensure that companies transparently handle AI-related layoffs and protect employees’ rights to information and consultation.
1. Why Do Bosses Regret AI-Driven Layoffs?
Many companies thought AI could solve all their problems, but they underestimated its limitations.
A survey by Careerminds of 600 HR professionals found that 75% of companies had implemented AI-based layoffs, but only 8.4% were satisfied with the results. 54.6% reported that the amount of human oversight needed by AI was much higher than anticipated. For example, Meta initially planned to lay off 60% of its staff; after the first round of layoffs, it discovered that while AI generated more code, the actual functionality was reduced, and accident rates increased by 40%. Employees had to spend 70% more time fixing errors, leading to the cancellation of the second round of layoffs. The CBA (Commonwealth Bank of Australia) cut 45 customer service positions, only to find that call volumes did not decrease, and the remaining employees had to work overtime, with supervisors also having to handle customer calls.
2. Rehiring Employees: A Victory That May Be a Loss
Rehiring employees does not necessarily mean an improvement for them.
For instance, CBA employees experienced weeks of anxiety due to financial concerns after being laid off, and even those who were rehired faced psychological distress. Klarna rehired customer service staff but changed their working arrangement to a “ride-hailing” model, hiring them on a freelance basis and treating them as cheap labor. One employee, Devrim, was rehired after a 40% pay cut. Meta’s situation was even worse: although AI generated 220% more code, the new features were of limited use to users, and accident rates increased, requiring employees to spend more time on troubleshooting. The positive feedback from employees decreased from 74% to 55%. In essence, company rehires are a recognition that AI cannot yet replace humans completely, not a sign of goodwill, and often result in worse conditions for employees.
3. Don’t Be Misled by “AI Transformation” – It May Be Just an Excuse for Layoffs
Some companies use AI as a cover for cost-cutting.
The research firm Forrester has termed this practice “AI washing” – companies attribute layoffs to AI when, in reality, they have not yet implemented a mature AI system to take over the tasks. Forrester predicts that by 2026, half of the layoffs attributed to AI will be reversed, but when rehiring occurs, the jobs may be moved overseas or the employees’ salaries will be reduced, making it unlikely for them to regain their previous benefits.
4. Europe Sets Rules for AI-Driven Layoffs
Europe does not prohibit AI-based layoffs, but it requires companies to be transparent.
Danish home service platform Hilfr and unions have agreed that when using AI to evaluate layoffs, companies must explain the data used and the weight given to each factor so that decisions can be legally reviewed. Insurance giant AXA reports on AI projects every six months and holds annual meetings to discuss the impact of AI on employees. The EU has revised regulations requiring multinational companies to notify employees in advance if they plan to lay off workers in multiple countries, allowing them to provide feedback and for management to respond. These rules are not designed to prevent layoffs but to ensure that companies do not use AI as a pretext and to give employees some say in the process, reducing the anxiety caused by sudden layoffs.
In Conclusion
AI can certainly perform certain tasks, but it is still far from being able to replace humans completely. Companies should not blindly implement layoffs based on AI, and employees should be wary of the guise of “AI transformation.” Europe’s approach serves as a reminder that layoffs are serious matters that require transparency and proper procedures.