If you record a radio show, teach adults their ABCs, or write the news for a living, it is time to reassess your career prospects in the future.
A fresh breakdown of United States government job projections has put a number on something workers in AI-exposed fields have feared to know. Not every job touched by artificial intelligence is safe, and some are shrinking fast.
The data, compiled from the US Bureau of Labor Statistics’ 2025–2035 employment projections, looks specifically at occupations rated “very high” for AI exposure that also require a bachelor’s degree but no prior work experience. In other words, the jobs fresh graduates walk into. Of those, ten are projected to lose ground over the next decade.
The Projected Shrinking List
● Adult basic education, adult secondary education and ESL instructors — the steepest projected decline of the group
● Broadcast announcers and radio disc jockeys
● Computer programmers
● News analysts, reporters and journalists
● Credit analysts
● Network and computer systems administrators
● Insurance underwriters
● Geographers
● Statistical assistants
● Graphic designers — the mildest decline on the list
At the top of that pile, literally reading and teaching for a living, adult and ESL educators carry the sharpest projected drop of any occupation in the “very high AI exposure” bracket.
Broadcast announcers and radio disc jockeys follow closely, alongside computer programmers and, notably, the very reporters and journalists writing stories like this one.
The Nuances
BLS carefully explains that “AI exposure indicates potential task transformation; not job loss.” This distinction matters as it notes that a “very high AI exposure” rating measures how much of an occupation’s day-to-day tasks could be reshaped by AI tools, such as drafting, summarising, coding, and transcribing.
It emphasizes that it is not a verdict that the job itself disappears. Some of these roles may shrink because AI absorbs entry-level tasks that used to require a full hire; others may shrink for reasons only loosely tied to AI at all, from shifting classroom funding for adult educators to the long decline of terrestrial radio.
The projections also cut both ways within the same exposure bracket. Data scientists, arguably as AI-exposed as any job on the list, show the strongest growth projection of the entire group, alongside logisticians, industrial and mechanical engineers, software developers, actuaries and financial examiners.
High AI exposure, in other words, is not destiny. It appears to matter less whether a job touches AI, and more whether the job is about building and directing the tools versus performing the routine, replicable tasks the tools are built to absorb.
Why this Matters to Students & Workers
The figures come from the American labour market, not Ghana’s or the wider region’s. But the occupations on the shrinking list travel well beyond US borders. This is because newsrooms, radio stations, credit desks and IT help lines look broadly similar wherever they sit, and the AI tools reshaping them are the same ones landing on desks across the board.
For young graduates weighing a first job in journalism, broadcasting, or entry-level IT support, it is worthy to note that the tasks that used to justify an entry-level hire in these fields are precisely the ones AI is learning to do first.
The Bureau’s own caveat stands that exposure is not the same as extinction. Whether these projected declines play out as fewer jobs, or simply fewer of the routine tasks within the same jobs, is the open question the next decade of hiring data will have to answer.
