IBM Triples Entry-Level Hiring as AI Eliminates Junior Roles

IBM is moving against the grain of a troubling industry trend, tripling its entry-level hiring even as artificial intelligence eliminates the routine work that traditionally filled those roles.

The countermove comes as new data reveals the scale of AI's impact on early-career employment. Stanford's Institute for Human-Centered AI reports that employment for software developers aged 22-25 has dropped nearly 20 percent since 2024, according to ADP payroll records covering millions of workers.

The decline extends beyond tech roles. Workers aged 22-25 in AI-exposed occupations including customer service and accounting saw a 13 percent relative employment decline. Meanwhile, workers aged 30 and over in the same fields saw headcount grow.

Why it matters

The immediate story is about youth unemployment, but the deeper risk is organizational. Companies that stop hiring junior talent are eliminating the apprenticeship layer where future leaders develop judgment through hands-on experience. That capability gap won't surface for years, when today's mid-level managers should be stepping into senior roles but lack the foundational skills that only come from doing the work.

The judgment problem AI creates

For decades, the path to leadership ran through grunt work. The analyst who spent two years in spreadsheets learned to spot numbers that didn't add up. The junior consultant who drafted 150 presentations before making partner absorbed what actually drives client decisions. The sales rep who knocked on doors learned to handle objections in real time.

These weren't just tasks to complete—they were the training ground for judgment that can't be taught in a classroom or absorbed from an AI output.

Now AI coding tools handle the boilerplate. AI drafts the first version of analysis. The tasks that used to fill a new hire's first year are being automated faster than organizations are redesigning their leadership pipelines.

What gets lost

The skills being automated away aren't just technical. They're the pattern recognition that comes from seeing hundreds of variations of the same problem. They're the instinct for when a client is truly concerned versus simply negotiating. They're the ability to recognize when data tells a story that contradicts the obvious conclusion.

These capabilities develop through repetition and real stakes. An AI tool can generate the work product, but it can't transfer the learning that comes from doing the work, making mistakes, and understanding why something failed.

Organizations cutting entry-level roles may see immediate efficiency gains. The cost shows up five to ten years later, when the pipeline of experienced mid-level talent who understand the business from the ground up simply isn't there.

IBM's decision to triple entry-level hiring suggests the company sees this risk clearly. While competitors automate away their apprenticeship layer, IBM is betting that maintaining the pathway from junior employee to seasoned leader is worth the investment, even when AI can handle many of the tasks those junior employees would traditionally perform.

The details were first reported by Inc., based on Stanford's 2026 AI Index data.