Mustafa Suleyman, CEO of Microsoft AI, has warned that artificial intelligence could fully automate most white-collar work within 12 to 18 months, a prediction that underscores the accelerating pace of AI development and its potential to reshape the global workforce.
In an interview with the Financial Times published Wednesday, Suleyman said Microsoft is pursuing “professional-grade AGI” capable of performing routine tasks for knowledge workers across nearly every industry.
”White-collar jobs—those sitting in front of computers, whether lawyers, accountants, project managers, or marketers—most of these tasks will be fully automated by AI within the next 12 to 18 months,” Suleyman told the Financial Times.
Suleyman’s remarks come as Microsoft accelerates its push for AI self-sufficiency following a renegotiated agreement with OpenAI that freed the company to independently pursue artificial general intelligence.
“We decided that this was a moment when we have to set about delivering on true AI self-sufficiency,” Suleyman said, signaling Microsoft’s intent to reduce its dependence on OpenAI by developing its own frontier models.
Microsoft recently formed an MAI Superintelligence team tasked with building what Suleyman calls “world-class, frontier-grade research capability in-house”. The company’s previous partnership with OpenAI had restricted Microsoft from pursuing AGI through 2030, according to Business Insider.
The warning arrives amid mounting evidence that AI is already disrupting knowledge work. Last month, Anthropic launched Claude Cowork, an autonomous AI agent capable of executing complex business workflows with minimal human oversight. The tool includes plugins covering legal, finance, sales, and data analysis tasks.
The launch triggered sharp declines in Indian IT services stocks, with shares of TCS, Infosys, and other software-as-a-service companies experiencing what analysts called a “SaaS-pocalypse” in early February. Investors fear the core advantage of traditional IT outsourcing firms is eroding as AI agents can perform similar work at a fraction of the cost.
Suleyman predicted that creating custom AI models will eventually become “as simple as making a podcast or writing a blog,” with intelligent agents capable of handling complex workflows within two to three years. He compared the coming shift to the way smartphones transformed daily life, suggesting AI agents will become indispensable collaborators for knowledge workers.
Economists have increasingly echoed these warnings. A Stanford Digital Economy Lab study found a 13 percent decline in entry-level hiring for jobs most exposed to AI since the rise of large language models. At the 2026 American Economic Association conference, researchers warned that the U.S. labor market has reached an “irreversible tipping point” from hiring to replacement.
