Every few months, the same alarm sounds. A prominent technology executive steps before a microphone, a journalist, or a Davos audience and announces that artificial intelligence is about to render entire professions obsolete. The offices are still open. The lawyers and accountants are still billing. And yet the warnings keep coming, each one more precisely calibrated for maximum urgency than the last.
The latest entry in this genre comes from Mustafa Suleyman, chief executive of Microsoft AI. In a conversation with the Financial Times, Suleyman predicted that AI will achieve human-level performance on most, if not all, professional tasks within 18 months. Lawyers, accountants, project managers, marketers, anyone who earns a living sitting at a computer, he said, should expect their work to be fully automated within that window. It is one of the sharpest timelines ever attached to office-work automation by a major technology leader, and it landed with predictable force.
It is also worth pausing to consider who is saying it, and why.
The pattern is the story
Suleyman’s warning did not arrive in a vacuum. It follows a now-familiar cycle that began in earnest in early 2025. Anthropic chief executive Dario Amodei warned that AI could wipe out half of all entry-level white-collar jobs, a prediction he has since walked back. Ford’s chief executive Jim Farley said AI would halve the number of white-collar positions in the United States. Elon Musk, speaking at Davos in January, suggested that artificial general intelligence could arrive this year, a claim that stretched credibility even by his own freewheeling standards of forecasting.
The pattern is not coincidental. The executives issuing these warnings are, without exception, the executives building and selling the products being warned about. It is a peculiar form of marketing, one that conflates ambition with inevitability and packages existential anxiety as product roadmap. The dynamic is not entirely unlike a pharmaceutical company insisting that everyone needs its drug, or an electric vehicle manufacturer declaring the end of the combustion engine on a timeline convenient to its order books. The prediction serves the predictor.
Musk’s track record on such pronouncements is instructive. Fully self-driving cars were two years away in 2016. They remain, by any conventional measure, unavailable. Neuralink was going to restore sight and mobility on schedules that kept moving. The Mars colony has been perpetually imminent. None of this invalidates the underlying technology. It does, however, suggest that attaching a precise timeline to a complex technological and societal shift is an act of performance as much as analysis.
What the evidence actually shows
The data, as opposed to the declarations, tells a more complicated story. A 2025 Thomson Reuters report examined how lawyers, accountants, and auditors are currently using AI.
The answer was document review and routine analysis. Productivity gains were described as marginal and well short of signalling mass displacement. More awkwardly, a study by the nonprofit Model Evaluation and Threat Research found that AI tools actually made certain software development tasks take 20 percent longer than they would have taken without AI assistance. This is a poor advertisement for imminent automation, particularly in a sector that has arguably adopted AI more enthusiastically than any other.
The economic returns from AI investment also remain heavily concentrated. Research by Torsten Slok, chief economist at Apollo Global Management, found that while profit margins at large technology companies surged by more than 20 percent in the fourth quarter of 2025, the broader Bloomberg 500 Index recorded almost no change. Wall Street, it seems, does not believe the disruption will spread far beyond the companies building it.
There are genuine signs of displacement. Employment consultancy Challenger, Gray and Christmas recorded approximately 49,000 job cuts attributed to AI so far this year. Microsoft itself cut 15,000 workers last year. These are real numbers and real people. But they are also a fraction of the millions forecast, and they sit uncomfortably next to Suleyman’s 18-month countdown.
Separating the noise from the signal
None of this means the transformation is not real. AI is changing knowledge work, compressing timelines, and automating tasks that once required sustained human effort. The question is not whether disruption is occurring, but whether it resembles the apocalyptic cliff edge that technology executives keep sketching for an audience of nervous shareholders and anxious professionals.
The honest answer is that nobody knows. The executives who issue these warnings are not lying, exactly. They are extrapolating from trajectories that may or may not hold, through adoption curves that may or may not compress, across organisations that have historically been slow to restructure around new technology. They are also, it must be said, the people with the most to gain from being believed.
Suleyman’s 18-month clock started ticking in February 2026. By August 2027, the offices will either be empty or they will not. History, and the evidence available today, suggests the latter is considerably more likely. The job apocalypse may be coming. It has just been perpetually 18 months away for most of the decade.
