TrajectoryGovernance & societyAI is starting to replace work, not just help with it

UpdatedyesterdayBeats6

AI is starting to replace work, not just help with it

The early framing of AI in the workplace was that it would augment workers, not replace them — a productivity tool, like the spreadsheet or the search engine. That framing has cracked. Specific roles, specific teams, specific functions are now being explicitly downsized with AI as the cited reason. The numbers are still small in aggregate, but the direction has changed and the question is no longer "if" but "how broad and how fast."

Timeline

  1. February 27, 2024

    Klarna's CEO publicly says generative AI is doing the work of 700 customer-service agents and the company has stopped hiring. The statement breaks the polite consensus that AI augments rather than replaces.

  2. September 4, 2024

    IBM announces a pause on hiring for roles that could be automated by AI — effectively a formal recognition that some jobs in its own workforce are now AI-substitutable.

  3. March 18, 2025

    Duolingo lays off contract translators and announces a shift to AI-generated content. Similar moves at smaller content and customer-service operations follow over the spring.

  4. August 5, 2025

    Goldman Sachs publishes research estimating that 25% of work tasks across developed economies could be automated by current AI — the first major bank to put a specific number on the displacement.

  5. December 2, 2025

    US Bureau of Labor Statistics shows entry-level white-collar hiring (legal, finance, accounting) down 18% year-on-year, with employer surveys citing AI as a factor for the first time.

  6. April 28, 2026

    Anthropic CEO Dario Amodei publicly predicts AI could eliminate half of all entry-level white-collar jobs in the next one to five years — the most explicit displacement claim from a frontier-lab leader to date.

Where things stand right now

AI displacement has moved from speculation to specific, named layoffs and hiring freezes. Customer service, copywriting, basic legal review, and entry-level coding are the visible front lines. The aggregate numbers are still small relative to the overall labour market, but the direction is clear and the major question is whether this stays in narrow pockets or broadens into a generational shift in how white-collar work is structured.