Microsoft AI chief says white-collar tasks automated within 18 months

If your job involves sitting at a computer all day—pushing spreadsheets, drafting briefs, or tracking project milestones—the head of Microsoft AI just slapped an 18-month countdown on your current routine.

Mustafa Suleyman, CEO of Microsoft AI, told the Financial Times that AI is set to hit human-level performance on most, if not all, professional tasks within the next 12 to 18 months. He specifically called out lawyers, accountants, project managers, and marketing pros, predicting that the vast majority of their day-to-day tasks will be fully automated in that window.

The coding canary in the coal mine

Suleyman stressed that this wave is not theoretical. He pointed directly to software engineering, where developers now rely on AI-assisted tools for the vast majority of their code production. In his words, that profound shift in how humans relate to the technology happened in just the last six months.

Yet, hyper-productivity comes with a brutal flip side. A recent Business Insider report highlighted growing "AI fatigue" among software engineers: the tools supercharged output, but also left workers completely burnt out as employers instantly raised expectations on how much work one person can juggle at once.

Suleyman is hardly a lone doomsayer among AI royalty. Anthropic CEO Dario Amodei previously warned that AI could wipe out 50% of entry-level white-collar jobs, arguing that builders have an obligation to be honest about what is coming. Computer scientist Stuart Russell went even further last year, cautioning that politicians could eventually stare down an 80% unemployment crisis spanning everyone from surgeons to corporate executives.

The alarm bells have already rung in Washington. Senator Bernie Sanders called Suleyman's prediction an "economic earthquake" and renewed calls for a moratorium on new AI data centers. Still, stopping that momentum looks nearly impossible: Big Tech poured hundreds of billions into data center infrastructure in 2025 alone, and President Donald Trump continues to back their construction.

Why it matters

For desk workers and knowledge creators, the takeaway is not distant sci-fi panic, but an aggressive 18-month deadline. If you are still using AI solely as an occasional email polisher, you are falling behind. Right now, testing how these agents dismantle your specific daily workflows is survival prep, not a hobby.

From an industry perspective, Microsoft is spending aggressively on enterprise products like Copilot while backing both OpenAI and Anthropic. Suleyman's commentary doubles as a wake-up call and a masterclass in enterprise sales: if high-level cognitive labor can be automated this fast, every enterprise executive feels pressured to buy in immediately or get left behind.

We think Suleyman's timeline leans ambitious when it comes to replacing entire job titles, given how slowly corporate bureaucracy, compliance, and legal liability actually move. However, at the granular task level, his prediction is eerily credible. The real immediate risk is not mass overnight pink slips, but the brutal "AI fatigue" already plaguing engineers spreading across marketing, finance, and legal teams, where fewer workers are expected to perform the output of whole departments.