AI Is Not Ending Work But It Is Rewriting the Career Ladder

The AI jobs debate is usually framed as a simple question: will machines replace people?

The data is starting to show something more complicated. AI is raising productivity, creating demand for new skills and changing how existing jobs are done. But the pressure is becoming much more visible at the start of the career ladder.

Stanford’s Digital Economy Lab updated its labour-market research this month using ADP payroll data through June 2026. It found no evidence of widespread, economy-wide job displacement. But among 22–25-year-olds in highly AI-exposed occupations, employment is around 19% below where it would have been if it had kept pace with less-exposed peers. The gap appears to be driven mainly by weaker hiring, not more people being fired.

That distinction matters. Layoffs make headlines. A graduate role that never gets opened does not.

What AI is actually changing

A lot of junior work has historically been “first version” work: drafting a memo, doing basic research, cleaning data, writing simple code, preparing documents or handling straightforward customer queries.

AI can now do parts of that work very quickly. For an experienced employee, that is useful leverage. They can get to a first draft faster and spend more time reviewing, challenging and making decisions.

The problem is what happens to the junior role underneath them.

PwC’s 2026 Global AI Jobs Barometer analysed more than one billion job adverts across six continents. In US entry-level roles, the jobs most exposed to AI were seven times more likely to ask for traditionally senior skills such as judgement and leadership. These “seniorised” entry-level roles have grown 35% since 2019, while other entry-level roles fell 10%.

That is a strange contradiction. Employers still want junior talent, but increasingly they want junior talent to arrive with skills that normally come from experience.

The productivity case is real

An NBER study of 5,179 customer-support agents found that access to a generative AI assistant increased issues resolved per hour by 14% on average. The gain was 34% for novice and lower-skilled workers.

A controlled GitHub experiment found developers using Copilot completed a specific coding task 55% faster than those without it.

So the productivity story is real. The harder question is where that productivity goes.

If AI lets a company produce more with the same workforce, that can support growth. If it lets the company produce the same amount with fewer people, hiring slows. If it helps create new products and services, new roles can appear.

The UK graduate market is already weak

Today’s UK data makes the entry-level issue difficult to ignore.

Adzuna says only 8,383 graduate jobs were advertised in July, down 45% from a year earlier and the lowest since it began tracking the data in 2016.

This cannot be blamed on AI alone. Higher employment costs, cautious corporate hiring and weakness across sectors such as healthcare, hospitality and logistics all matter.

But the backdrop is uncomfortable. The latest ONS estimate shows 1.012 million people aged 16-24 were not in education, employment or training in January to March 2026, equivalent to 13.5% of that age group.

India shows a different version of the shift. Naukri’s July JobSpeak data showed overall IT hiring returning to 6% year-on-year growth, while AI and machine-learning roles grew 33%. Fresher hiring also grew 6%, a useful reminder that this is not a one-way story of disappearing jobs. Demand is moving towards different skills.

The risk is the missing training ground

Technology has changed jobs before. The useful lesson is not that technology always creates more jobs, or that it always destroys them. It is that tasks move.

Economists Daron Acemoglu and Pascual Restrepo describe this as a balance between displacement and reinstatement. Automation removes human tasks; new tasks can rebuild demand for labour.

What feels different with generative AI is that it can automate cognitive work that used to double as training.

You learn judgement partly by doing the basic work, getting corrected, seeing patterns and repeating the process. If firms remove too much of that layer, they may save time today but create a talent problem later.

That is the part I think deserves more attention. The biggest risk is not that every junior job disappears. It is that we make entry-level jobs harder to enter while still expecting the same pipeline of experienced workers five or ten years from now.

What needs to change

The answer is not to preserve pointless busywork.

Companies need to redesign junior roles around AI: give people the tools, but keep structured review, coaching, apprenticeships and real responsibility. Universities need to teach AI fluency alongside writing, numerical reasoning, source-checking, communication and domain knowledge.

For young workers, simply knowing how to use AI will not be enough. The advantage will come from knowing when the output is wrong, what context it is missing and how to turn it into something useful.

AI is not ending work. But it is changing how people become good at work.

The companies and economies that handle this best will not be the ones that automate the fastest. They will be the ones that work out how to keep building human capability while they do it.

Sources: Stanford Digital Economy Lab, Canaries in the Coal Mine?, revised 12 August 2026; PwC, 2026 Global AI Jobs Barometer, 15 June 2026; Office for National Statistics, UK NEET statistics, 28 May 2026; Adzuna UK job-market data reported by The Guardian, 24 August 2026; Naukri JobSpeak, July 2026; Brynjolfsson, Li and Raymond, NBER, Generative AI at Work; GitHub productivity research; Acemoglu and Restrepo, NBER, Automation and New Tasks.

Disclaimer: This article is for educational purposes only and should not be considered financial advice. Always conduct your own research before making investment decisions.

Article uses AI for refining words