Microsoft to Lay Off Thousands of Employees

Microsoft is preparing another round of layoffs, with nearly 5,500 employees expected to lose their jobs next week as the company ramps up investments in artificial intelligence infrastructure.

According to reports from Business Insider, the latest workforce reduction will primarily affect employees in sales and consulting roles. Previously anticipated job cuts within Microsoft’s Xbox gaming division are also expected to be included. In total, fewer than 5,500 employees will be impacted — representing less than 2.5 percent of Microsoft’s global workforce of approximately 220,000 people. Some affected employees are expected to be offered alternative positions within the company.

The upcoming layoffs follow several previous workforce reductions. Microsoft eliminated 6,000 positions in 2025, followed by approximately 300 job cuts at its subsidiary LinkedIn. In 2023, the company laid off 10,000 employees.

Microsoft Shifts Focus to AI Infrastructure

The ongoing workforce reductions are part of Microsoft’s broader strategic realignment. The company is reducing costs in traditional business areas to free up capital for large-scale AI investments. Since 2025, Microsoft has committed approximately USD 190 billion to expanding its AI capabilities.

A significant portion of that funding is being directed toward new data center construction. Microsoft currently operates around 300 data centers across 34 countries. Among the company’s latest AI initiatives is the AI assistant Scout, which is built on the open-source OpenClaw platform. However, before the planned expansion of the platform, security researchers identified five zero-day vulnerabilities in the OpenClaw system.

Developers Push Back Against Token-Based Pricing

Alongside external AI platforms, Microsoft is also developing its own AI models, including those designed to enhance GitHub Copilot. However, the launch of the Copilot Pro subscription, priced at USD 39 per month, has drawn criticism from developers due to its consumption-based billing model.

Developers argue that the monthly token allowance can be exhausted far too quickly during intensive use. According to reports, one example showed that roughly eight percent of the monthly token quota was consumed within just two hours.

(ll)