AI is reshaping entry‑level work
New reporting says white‑collar jobs — including many junior roles — are more exposed to automation than people realise, and Gen Z is already shifting toward multiple part‑time gigs rather than single full‑time careers. That combination means employers may increasingly value evidence of judgement, cross‑functional skills and the ability to work with AI tools, not just task execution. (fortune.com)
A new split is opening in the job market: the work most exposed to artificial intelligence is often the work companies used to give beginners, while many younger workers are already piecing together income from several jobs instead of one career-track role. (fortune.com) Deputy said on April 6 that poly-employment has reached a decade high in U.S. shift work, and Gen Z now makes up 55% of workers holding multiple jobs at once. The same report says Gen Z has overtaken Millennials in frontline work and now accounts for 41% of shift workers. (news.deputy.com, deputy.com) That trend started in restaurants, retail, care, and other hourly jobs, but it lines up with a second change happening in offices: software is getting good at the exact kind of repeatable work junior staff used to do first. Think first-draft research, formatting slides, summarizing meetings, cleaning spreadsheets, or writing boilerplate code. (fortune.com, anthropic.com) Anthropic’s March 24 Economic Index says about 49% of jobs have already seen at least a quarter of their tasks performed with Claude. Its January report found that tasks needing a college degree were sped up about 12 times in Claude.ai, versus about 9 times for tasks needing a high school education. (anthropic.com, anthropic.com) That does not mean half of all jobs disappear. It means the first layer of work inside many jobs can be done faster or handled automatically, which is exactly where companies used to train new hires by giving them smaller pieces of the job. (anthropic.com, fortune.com) The International Labour Organization reached a similar conclusion in a 2025 working paper: generative artificial intelligence exposure is highest in clerical and administrative work, and the risk is much more about tasks being transformed than whole occupations being erased overnight. (econstor.eu) So the old bargain for graduates gets shakier. A company that once hired a junior analyst to turn raw information into a draft may now want one person who can prompt an artificial intelligence tool, check the output, spot what is wrong, and explain the result to sales, product, or legal teams. (fortune.com, anthropic.com) Anthropic’s newest report also found that experienced users get better results than newer users and are more likely to attempt higher-value tasks. In plain English, knowing how to work with the tool is starting to look less like a bonus skill and more like basic office literacy. (anthropic.com) That helps explain why employers are putting more weight on judgment than on output alone. If a chatbot can produce the first draft in 30 seconds, the scarce skill is deciding whether the draft is accurate, useful, on-brand, lawful, and worth sending. (fortune.com, anthropic.com) It also helps explain the rise of patchwork careers. If full-time entry points narrow in white-collar fields while flexible work expands elsewhere, more workers in their early 20s may build income like a playlist instead of a ladder: one part-time role, one contract project, one freelance stream, all running at once. (news.deputy.com, fortune.com) The awkward part is that companies still need future managers, but the traditional training ground for future managers was entry-level work. If software takes too much of that rung away, firms may save money now and then discover later that they trained too few people to make hard calls when the software runs out of road. (cnbc.com, fortune.com)
