Junior Tech Hiring in Europe Plunges 73.4% in a Year

A one-year fall of that size leaves the intake at 26.6% of its previous level ^[13]. That figure carries the whole claim that AI is removing entry-level technical work. The column reports the 73.4% drop without naming a dataset or the twelve months it covers ^[18]. Entrepreneur notes that its contributors' opinions are their own ^[15].

The best-documented case in the same piece runs the other way. Ford put around 900 AI cameras on quality control and found the automated systems repeatedly missed subtle defects that experienced human inspectors caught in seconds ^[6]. Those inspectors got fast enough to beat a camera by spending years on a line. The company brought back 300 experienced engineers to overhaul the tools and manage complex cases ^[7].

The trade-off is between two costs that land in different quarters. Rewriting requisitions around responsibilities and expected outcomes instead of task lists, as the article recommends, costs a quarter of process work and some discipline in the interview loop ^[12]. Deferring the junior intake costs nothing this quarter either. That cost arrives when senior staff need replacing and nobody junior was ever hired to replace them.

The founder data is weaker support than it first looks, because it measures how companies get started and not how they staff up afterwards. Carta's share of single-founder US startups went from 23.7% in 2019 to 36.3% in mid-2025 ^[4], a gain of 12.6 points, or about 53% in relative terms ^[14]. The article treats AI as an important factor, on the grounds that one person can now research, build, launch and maintain a product without hiring a specialist for each role ^[16].

This may be an ordinary hiring downturn wearing an AI label, and the record here cannot refute that. What survives either reading is the screening problem: one interview loop cannot both suppress AI use and test AI fluency. Gartner projects that most companies will test the skills of future employees AI-free, because heavy GenAI use is degrading critical thinking ^[5]. LinkedIn counts a sixfold jump in mentions of AI literacy in job postings in a single year ^[9], and Zapier has set a bare minimum AI fluency expectation for all new hires ^[10].

Job architecture is the cheaper half of the response. Airtable's chief executive Howie Liu restructured the company around AI after an unreliable post claiming 'Airtable is dead' went viral, creating a product team for rapid AI experiments and an infrastructure team for slower strategic work, and went back to writing code himself ^[8]. At Udora, a product manager built a functional internal admin interface with AI tools; it was not polished, it launched in days, and it took no development time ^[11]. The article's list of calls that should stay human is short and specific: disputes, legal and privacy risk, financial issues, brand image moments ^[17].