Microsoft Pushes Internal AI Adoption: Managers Told to Factor Copilot Use Into Employee Reviews

Microsoft is taking a harder stance on employee AI adoption, instructing some managers to factor the use of internal AI tools—particularly GitHub Copilot—into employee performance reviews.

The push, first reported by Business Insider, is part of a broader effort to drive internal uptake of the company’s flagship AI products amid growing signs that even Microsoft’s own employees are opting for alternatives like ChatGPT.

In a memo seen by BI, Julia Liuson, President of Microsoft’s Developer Division, told managers that AI usage should now be considered a “fundamental part of how we work,” equating it with key performance traits such as collaboration, data-driven decision-making, and communication.

“Using AI is no longer optional — it’s core to every role and every level,” Liuson wrote.

Managers were instructed to evaluate how effectively employees integrate internal AI tools, such as GitHub Copilot, into their daily work as part of “holistic” performance evaluations.

Growing Internal Discontent with Copilot

The policy comes amid increasing signs of dissatisfaction with GitHub Copilot, even across parts of Microsoft’s own workforce. Despite being the company’s signature AI coding assistant, Copilot is reportedly being shunned by some employees, who prefer using OpenAI’s ChatGPT, which they consider to be more versatile and accurate.

Internal feedback, according to people familiar with the matter, shows that developers often find ChatGPT more effective for writing, debugging, and explaining code—despite Copilot being tailored specifically for those tasks. Some developers have also raised concerns about Copilot’s inconsistent performance, limited context handling, and integration issues with certain programming environments.

The situation is awkward for Microsoft, which has invested billions into OpenAI and co-developed Copilot as a commercial application of GPT technology. ChatGPT, on the other hand, is OpenAI’s direct-to-consumer product—and it’s pulling users away from Microsoft’s own platform.

The problem is compounded by rising competition in the AI developer tool space. Rivals like Cursor, which builds on top of GPT-4 and offers deeper integration for code workflows, have gained traction among professional developers. A recent Barclays note cited data suggesting Cursor had surpassed Copilot in usage among some developer groups.

Meanwhile, Microsoft has allowed employees to use some approved third-party tools such as Replit, which has gained popularity for its ease of use and real-time feedback features. Yet even with this flexibility, internal preference for ChatGPT remains strong.

The growing reliance on OpenAI’s flagship chatbot rather than Microsoft’s branded tools adds complexity to the companies’ evolving partnership. Microsoft is reportedly renegotiating its deal with OpenAI, and internal documents show that disagreements have emerged around intellectual property and access rights—especially as OpenAI weighs acquiring Cursor competitor Windsurf.

According to sources, OpenAI and Windsurf want to block Microsoft from inheriting IP access rights through its OpenAI agreement. These discussions highlight how Microsoft’s aggressive Copilot strategy and internal adoption struggles are intersecting with broader tensions in the AI industry.

A Cultural Shift with Potential Blowback

Microsoft’s effort to mandate AI usage reflects a cultural shift within the company, aiming to align its workforce with the same AI-first vision it sells to enterprise clients. But turning AI use into a performance metric also risks internal pushback, especially if employees feel forced to adopt tools they view as less effective than third-party alternatives.

However, Microsoft seems determined to push forward. While some teams are informally integrating AI-use metrics into reviews, others are considering formalizing the requirement starting in the next fiscal year.

The company’s leadership sees this as vital not just for productivity, but also for product refinement—believing that widespread internal use of tools like Copilot will help close feature gaps and improve competitiveness.