Artificial Intelligence Will Not Replace Workers but Will Assist Them: Microsoft’s AI Chief Changes His Mind

Mustafa Suleyman

Mustafa Suleyman, the head of AI at Microsoft, sparked discussion a few months ago with a series of statements. The manager had predicted that within 12-18 months, there would be a significant impact on jobs, particularly among so-called 'white-collar' workers. Essentially, according to Suleyman, artificial intelligence would eliminate many jobs, focusing particularly on intellectual or administrative tasks. However, it seems the manager has changed his mind and now sees AI differently.

AI as an Assistant

In a surprising turnaround, Mustafa Suleyman now has a new perspective on artificial intelligence, emphasizing that he has been misinterpreted in the past. During the Decoder podcast by The Verge, the manager defined AI as an assistant and not as a tool meant to replace human jobs. Many less critical office tasks will become "digitized and automated." Thus, AI will replace "tasks" and not "jobs." This important distinction helps the Microsoft manager clarify his view on the future of AI. An agentic AI can therefore assist and support workers by taking on the more repetitive and time-consuming tasks.

It is worth noting that Mustafa Suleyman joined Microsoft in 2024, aiming to lead the Microsoft AI division and manage various products, including Copilot. Previously, he worked for DeepMind, Google/Alphabet's cutting-edge AI company.

AI and Work: Insights from Brad Smith

Brad Smith, Vice Chair and President of Microsoft, also discusses AI and work in a lengthy blog post on the company's official blog that is worth reading carefully—not just for those in tech but for anyone apprehensive about the advancement of artificial intelligence.

The starting point is unusual: this year's American graduation ceremonies, where several graduates openly booed at the mere mention of AI. A signal that Smith does not dismiss as mere generational Luddism; instead, he sees it as a wake-up call that the industry must respond to.

Smith does not mince words: young people entering the job market today face a perfect storm. Automation is already eroding entry-level positions, tech companies are cutting positions to fund massive investments in AI infrastructure, and on top of this, there are geopolitical tensions and post-pandemic corrections. Ignoring these concerns would be a mistake.

Smith's central thesis is that technology has never, in the long run, destroyed jobs—it has transformed them. He cites the example of photography, which in 1838 was heralded as the "death of painting"; instead, it stimulated new artistic movements like Impressionism and Cubism. The same goes for spreadsheets, which did not eliminate accountants but enabled them to build infinitely more sophisticated financial models.

For individual workers, Smith suggests stopping thinking of their work as a title and starting to view it as a set of tasks. Which tasks can AI do on its own? Which can I do better with AI? Which require exclusively human judgment? Answers to these questions indicate where to invest in oneself.

Soft skills—curiosity, creativity, communication, compassion, courage—become even more crucial differentiators.

For companies, the message is equally clear: use AI to accelerate internal learning, not to replace it. Above all, build proprietary AI capabilities while controlling your own data—giving up your know-how to other models means funding tomorrow’s competition.

Smith concludes by calling for a broader public conversation that includes unions, universities, governments, and civil society. It’s not enough for tech companies alone to decide the pace and direction of transformation. A post that, regardless of its source, raises questions that anyone working with technology should consider.