MIT Report Finds That AI Isn't Killing Jobs, It's Reorganizing Them; But Entry-Level Roles Are Taking the Hit

A new MIT Technology Review report is challenging the growing belief that artificial intelligence is rapidly destroying jobs. Instead, it finds that AI is mostly reshaping how work is organised inside companies, changing tasks rather than wiping out entire roles.

But while the overall job market remains stable, the research exposes a clear pressure point: younger, entry-level workers are feeling the impact first.

The study examined US labour market data and found no strong evidence that AI has caused widespread unemployment so far. Even in sectors considered highly exposed to automation such as software development, finance, and customer service employment levels have not collapsed.

In fact, in some cases, unemployment rates in these roles are lower than in less AI-exposed jobs.

The findings come as the public raises fears about whether tools like ChatGPT and other generative AI systems will replace white-collar work. Large-scale layoffs in parts of the tech industry have added to those fears. But MIT researchers argue that, at least for now, the data tells a more restrained story.

AI Is Reshaping Work Rather Than Removing Jobs

The MIT analysis is a simple conclusion: AI is reorganising jobs, not eliminating them.

Companies are increasingly using AI tools to handle specific tasks such as drafting text, summarising information, analysing data, and responding to customer queries. However, instead of replacing workers outright, many firms are redistributing work. Employees are shifting away from repetitive tasks and spending more time overseeing, checking, or refining AI-generated output.

Researchers describe this as 'task reshaping' rather than job destruction. It reflects a pattern seen in earlier technological shifts, such as the introduction of computers and the internet. Those technologies did not immediately eliminate large numbers of jobs either, but they did significantly change what people actually did within their roles.

One important takeaway from the study is timing.

Economists stress that even major technologies usually take years, sometimes decades, to fully reshape labour markets. Businesses must first adopt the tools, redesign workflows, and only then adjust hiring patterns.

For now, there is no sign of a dramatic, economy-wide shock caused by AI. Instead, the effects are uneven and still developing.

Entry-Level Workers Are Facing The Sharpest Changes

Where the study does see a noticeable shift is at the bottom of the job ladder.

Entry-level roles in AI-exposed industries, particularly software engineering, customer support, and data-heavy administrative work are becoming more difficult to access. These roles often rely on structured, predictable tasks, which are precisely the kinds of tasks modern AI systems are becoming good at performing.

This does not necessarily mean those jobs are disappearing entirely.

Instead, companies appear to be hiring fewer junior employees while keeping experienced staff. In some cases, they are also redesigning entry-level roles so they involve more oversight of AI systems and less routine work.

This may be linked to a deeper divide in how knowledge is used at work. Entry-level employees often rely on 'codified knowledge': rules, training, and patterns that can be learned quickly and replicated. AI is increasingly capable of handling these tasks.

More experienced workers, however, rely on 'tacit knowledge,' which are judgement, intuition, and context built up over years. That kind of knowledge is far harder for AI to replicate, which may explain why senior roles are currently more stable.

A Labour Market In Transition, Not Collapse

The MIT study also points out that AI adoption across companies is still uneven.

Many firms are experimenting with the technology rather than fully integrating it into core operations. This means its impact on jobs is still limited and difficult to measure precisely.

Some studies have found that employment among younger workers in highly automatable roles has fallen since the rise of generative AI, while older workers in the same fields have been less affected or even seen growth.

However, economists caution that it is too early to draw firm conclusions. The current phase may represent an adjustment period rather than a permanent shift.

There is also evidence that AI is improving productivity in some areas, even if the gains are not yet large enough to transform the economy. In practice, this means workers may be doing more with the help of AI tools, rather than being replaced by them.

Still, the study warns that entry-level disruption should not be ignored. If hiring pipelines weaken over time, it could reshape how people start careers, especially in industries like tech, where junior roles traditionally serve as the foundation for progression.