AI begins to reshape entry-level work

AI companies are making ambitious claims about tools that can automate or augment human work, while major employers are investing heavily in systems that could reduce hiring needs. Investors are increasingly asking whether new roles should be filled by people or by AI agents designed to perform specific tasks, including some skilled work.

Large language models have rapidly improved in software-related benchmarks. Three years ago, they could reliably handle tasks that took humans seconds or minutes, while current models can complete more complex work taking an hour or so. Similar patterns are emerging at earlier stages in financial analysis, early legal work and some entry-level creative jobs.

Stanford University’s analysis of wage and jobs data finds employment among 22 to 25-year-olds has fallen 2.7% since ChatGPT became widespread, rising to 12.8% in the most AI-exposed sectors such as finance, software and creative industries. Other economists argue that factors such as interest rate rises may also explain the shift, but OECD jobs posting data shows exposed sectors weakening in several markets, with the UK notably vulnerable because of its service-sector concentration.

AI usage has surged in 2026 as companies track token consumption and push employees to find productivity gains from advanced models. Heavy use of agentic systems has produced large bills, prompting some firms to ration access, while cheaper models derived from freely available Chinese systems may lower costs and broaden adoption.