Bosses Replace Staff with AI to Cut Costs, But End Up Spending More

Many companies were initially enthusiastic about replacing or reducing their human workforce with artificial intelligence, hoping to cut long-term operational costs. However, a harsh reality is now emerging: AI usage bills are swelling and often exceeding the salaries of the employees who were laid off. A recent report from accounting firm KPMG, released in early July 2026, has opened the eyes of corporate leaders to the fact that AI is not a cheap, free-to-use solution. During trial periods, AI service providers typically offer fixed-price contracts or large discounts to entice companies. Yet, as computing power costs have soared, almost all large language model developers have now switched to payment systems based on usage or per unit of data processed. This means the more AI is used, the higher the bill. Many directors who initially thought costs would be fixed are now shocked to receive invoices that keep rising dramatically. The KPMG survey, which involved 2,145 senior executives from 20 countries, found that 29 per cent admitted they did not know exactly where the increased AI costs were coming from, while 33 per cent stated their lack of understanding of AI economics was a hindrance. Only about a quarter of companies had a clear and measurable system for overseeing AI budgets. The initial logic for many CEOs was simple: replace employees with AI to eliminate costs for salaries, benefits, and leave. The reality has proven different. AI requires expensive server infrastructure and electricity, specialist staff, and still needs human intervention to correct its mistakes. As a result, AI usage bills, measured in tokens, have multiplied because usage was not capped. In some reported cases, companies ended up spending two to three times more on AI than the budget for the teams they had dismissed.