Mustafa Suleyman, the head of artificial intelligence at Microsoft, has issued a stark warning that artificial intelligence could soon take over a large share of white-collar work, including roles traditionally seen as highly skilled.

Speaking in an interview with the Financial Times, Suleyman said Microsoft is seeking to deepen its presence in the enterprise sector by developing what he described as “professional-grade AGI”. According to him, this would be an advanced form of AI capable of handling many of the same tasks performed by human professionals, allowing businesses to deploy AI tools for everyday knowledge work.

Suleyman suggested the impact on the workforce could arrive sooner than many expect. He said that anyone whose job primarily involves working on a computer could be affected. Roles such as lawyers, accountants, project managers and marketing professionals, he explained, rely on tasks that AI systems are rapidly learning to perform. In his view, most of these activities could be automated within the next 12 to 18 months.

His comments come as companies across industries accelerate their adoption of AI tools. Recently, Anthropic’s Claude Cowork model unsettled markets after raising concerns about the long-term prospects of SaaS firms, including India-based IT majors Infosys and TCS.

Suleyman also believes building AI models will soon become far more accessible. He compared the process to creating everyday digital content, suggesting that designing a new AI system could one day be as easy as writing a blog post or producing a podcast. This shift, he said, would make it possible to tailor AI systems for specific organisations and even individuals.

Looking further ahead, Suleyman said AI agents could become far better at managing complex workflows in large institutions within the next two to three years.

He also indicated that Microsoft plans to strengthen its in-house AI capabilities. Following a revised agreement with OpenAI, the company aims to reduce its dependence on external partners. Suleyman said this move reflects Microsoft’s push towards long-term AI self-sufficiency, hinting that new AI models could be unveiled in 2026.