Terapeut Ocupațional
10AI USED FOR10%moving to54%
THE DOOR, NOWB
THE DOOR, 20 YRSC
STARTING PAY£25,000 - £30,000
REGULATORY MOATThe law requires human oversight. That human could be you.
CAREERMASH (RO)FROM SUBJECTS ALLIED TO MEDICINE →
Terapeut Ocupațional
Terapia ocupațională este o carieră dinamică și satisfăcătoare care se concentrează pe ajutarea indivizilor să își recâștige independența și să îmbunătățească calitatea vieții. În România, terapeuții ocupaționali lucrează în diverse medii, inclusiv spitale, școli și servicii de sănătate comunitară, având un impact semnificativ asupra vieții pacienților lor. Cu o cerere în creștere pentru servicii de sănătate mintală și reabilitare, această profesie este esențială și oferă o cale de carieră satisfăcătoare pentru cei pasionați de ajutorarea altora.
This job holds up well against AI
AI is used for very little of what this job actually involves today, and the law requires human oversight. That human could be you. That is about as safe as careers get right now, though it is still worth learning to use the tools well.
How this job changes over time
Our best estimates, shown as ranges and grades, not exact predictions.
| WHAT CHANGES | Now | 5 yrs | 10 yrs | 20 yrs |
|---|---|---|---|---|
| Tasks AI is used for | 14% | 27% | 42% | 54% |
| Number of jobs | 98-105% | 95-105% | 88-105% | 72-95% |
| How hard to get in | B - achievable | B - achievable | B - achievable | C - hard |
| Job security | Strong | Strong | Holding | Law's call |
Right now, AI is already used for about 14% of the day-to-day work in this job, and by 20 years from now that could be around 54%. There are likely to be a similar number of of these jobs over time, very roughly 72-95% of the 2024 number, 20 years out. Getting your first job here is fairly easy today, and it looks set to get harder. The law requires human oversight. That human could be you.
What happens, and when
WITHIN 5 yrs
Strong
At this horizon the model has AI used for about 27% of this job's day-to-day work, and the number of these roles is projected at roughly 95-105% of the 2024 total. Getting in is graded B - achievable, and job security here is rated Strong.
WITHIN 10 yrs
Holding
At this horizon the model has AI used for about 42% of this job's day-to-day work, and the number of these roles is projected at roughly 88-105% of the 2024 total. Getting in is graded B - achievable, and job security here is rated Holding.
WITHIN 20 yrs
Law's call
At this horizon the model has AI used for about 54% of this job's day-to-day work, and the number of these roles is projected at roughly 72-95% of the 2024 total. Getting in is graded C - hard, and job security here is rated Law's call.
The honest bottom line: AI will keep taking on more of the routine output here, but the law requires human oversight. That human could be you. That is what keeps this job valuable once the easy work moves to software. Get fluent with the tools, and make sure you are the one holding the part that cannot be automated away.
How to aim for this career
1
Head for the licence
A professional qualification or registration is the real protection here. Find out which body governs this career and start working towards what it requires.
2
Get comfortable being accountable
Practise explaining decisions clearly and owning the outcome. Being the person who signs off is the actual job.
3
Use AI as a second opinion
AI can draft and check work, but someone has to be legally responsible for it. Build the judgement to catch what it gets wrong.
Careers that use similar skills, and hold
Worth a look if you like the sound of this path. Each one has something this job does not: a moat.
Sources: exposure blends published AI-usage research (Anthropic 2026, observed real-world usage; OpenAI 2026) with our own UK reviews of labour-market moats, robotics evolution and scientific developments. Job-security forecast from OpenAI, "The AI Jobs Transition Framework" (Richmond, 2026, OpenAI Economic Research), CC BY 4.0. Grades and verdicts are Careermash editorial judgment: we show forecasts as forecasts and own our conclusions.