AI and the future of Health and Science work: what changes, what stays human

Key facts

• On the ILO's 2025 index, nursing professionals score 0.25 and health care assistants 0.14 out of 1, in its lowest band, "not exposed". Medical records technicians score 0.52. These are global generative-AI task-exposure estimates; not a Malaysian forecast; not a job-security rating^[1].
• Healthcare employers in the World Economic Forum (WEF) survey said people mainly do 50% of tasks today and expected 34% by 2030, with 35% done by people and technology together. Global employer expectation, published 2025^[2].
• The same survey put nursing professionals among the largest-growing jobs to 2030, partly because populations are ageing. Global employer expectation, published 2025; not a Malaysian shortage figure^[2].
• Historical example (May 2025, news report): the Health Ministry began piloting AI-supported chest X-ray screening for lung disease in seven government clinics, with a low-dose CT scan to confirm what the AI flags^[3].

Health and science work mixes hands-on care with a lot of reading, recording and checking. In the ILO's index, the record-keeping health occupations score far higher than the bedside ones. Nobody has measured what that means for Malaysian clinics yet, so this page gives you questions to test, not answers.

For routes in, SPM subjects, costs and the registration some health jobs need, start with Health & Science after SPM.

How to read this page

Observed means someone measured it. Projected means someone estimated the future, so we say who, when and for where. Projections are often wrong.

The scores come from the International Labour Organization (ILO). They estimate how much of an occupation's current tasks generative AI (tools that write, summarise or make images from a prompt) could potentially do, from 0 to 1^[1]. "Not exposed" is a technical label: most of the occupation's tasks scored consistently low for what generative AI can do today^[1]. It does not mean AI will never change the work, and it says nothing about robots, cost or Malaysian employers. Exposure is not job loss either: the ILO says that depends on adoption and on whether workers get the chance to learn the tools^[1].

What is already happening (observed)

• Malaysia, historical (May 2025). The Health Ministry's Lung Health Initiative 2025-2030 started an AI-supported chest X-ray pilot in seven government clinics in May 2025, to help find small nodules that ordinary X-rays can miss. Flagged patients have a low-dose CT scan to confirm^[3]. This comes from a news report of the launch. We could not confirm the pilot's current scale or results on a Ministry page today. The AI helps decide who needs a closer look; clinicians and further tests decide what is wrong.
• One US usage dataset. In Anthropic's February 2025 data (Claude.ai conversations sorted into US job categories), the lowest-paid and the very highest-paid occupations showed very low use of its AI. These were generally jobs needing a lot of manual skill, such as obstetricians^[4]. That is usage data for one AI product. It does not show Malaysian demand or future job security.

What people project, and who says so

• Healthcare employers worldwide. In the WEF survey, 91% of medical and healthcare employers expected AI to transform their business by 2030, and 63% expected the same of biotechnology. They said people mainly do 50% of tasks today, technology 23% and the two together 27%, and expected 34%, 31% and 35% by 2030. Global employer expectation, published 2025^[2].
• Working with AI more than replaced by it, in this sector. The WEF says nearly half of the expected fall in tasks done by people alone in healthcare and government is expected to come from people working with technology, not from automation. Global employer expectation, published 2025^[2].
• Care jobs. WEF employers put nursing professionals and personal care aides among the largest-growing jobs to 2030, and the WEF links nursing growth partly to ageing populations. Global employer expectation, published 2025^[2].
• Data skills inside health. Data analysts and scientists, and AI and machine learning specialists, topped healthcare employers' growing roles, while administrative assistants were shrinking. Global employer expectation, published 2025^[2].
• Robots, a US model. Anthropic's September 2026 model rates nursing as low exposure to robots and healthcare support as relatively low^[5]. It is US model data, not Malaysian demand.

For Malaysia: we did not find a current Malaysia-specific study of AI and health jobs. TalentCorp's 10-sector study, reported in August 2025, covered medical devices and pharmaceutical manufacturing but not health services^[6]. Local demand, registration, pay and public-sector hiring need Malaysian sources: the Health & Science guide is the place to start.

Task by task

These cards are our reading of the sources for a classroom exercise, not a forecast. A practitioner may disagree, and their reasons are worth hearing.

Examine a patient and explain a diagnosis

• What a person does: takes the history, examines, decides and answers for the decision.
• What AI might assist: notes, letters and looking things up.
• What this source does not show: whether Malaysian clinics will adopt these tools, or how fast.

Care for patients on a ward

• What a person does: gives medicines, watches for change, helps people wash, eat and move.
• What AI might assist: charting and handover notes.
• What this source does not show: staffing levels, shift patterns or robots on wards.

Screen a chest X-ray

• What a person does: takes the image, reviews what is flagged, arranges further tests.
• What AI might assist: flagging images that need a closer look.
• What this source does not show: how accurate the tool is in Malaysian clinics, or whether the pilot grew.

Run lab tests on samples

• What a person does: handles samples safely and spots results that look wrong.
• What AI might assist: checking results and drafting reports.
• What this source does not show: how much Malaysian labs already automate.

Code and file patient records

• What a person does: checks records are accurate and kept private.
• What AI might assist: coding, filing and booking, which are structured and repetitive.
• What this source does not show: whether these jobs shrink, change or move in Malaysia.

Analyse public health data

• What a person does: decides what matters for a community and explains it.
• What AI might assist: running the numbers and drafting charts.
• What this source does not show: which Malaysian employers hire for this, or at what level.

New tasks that are appearing

• Checking what an AI tool flagged and explaining it to a worried patient.
• Running and auditing AI screening: who gets screened, what happens next, what the tool misses.
• Keeping health data accurate and private. Privacy and data security is one of the seven principles in Malaysia's voluntary National Guidelines on AI Governance and Ethics (MOSTI, September 2024)^[7].
• Health data analysis, which healthcare employers list among growing roles (global employer expectation, published 2025)^[2].

This list is our reading of the sources above, not an official one.

What stays human: hypotheses to test

These are practitioner hypotheses to test, not findings. The sources do not show that any of these tasks will stay unchanged.

• Does this work need touch, presence and trust at the bedside?
• Who examines the patient and answers for the decision?
• Who explains bad news to a family, in their language?
• Who notices that something is wrong before any number shows it?

Ask a practitioner which of these hold in their job, and which have already changed.

What this means for you

• Do not read "not exposed" as "safe". It is a global estimate for generative AI only. Choose health work because the work suits you, not because a score is low.
• Add data skills. Even in clinical work you will read dashboards and AI outputs. Basic statistics will help you judge them.
• Protect patient data from day one. Never paste real names, IC numbers or health details into a public AI tool, even for homework.
• Look beyond doctor and nurse. Lab science, pharmacy, physiotherapy, imaging, records and public health have different mixes of hands-on and screen work, and different routes in.
• Check the route before the trend. Entry requirements, registration, shift work and costs matter more to your next step than any AI forecast. The Health & Science guide covers them.

Your next step

1. On Day 2, choose the Health and Science taster (a practice public-health case with no real patient data) or, if your cohort runs four labs, the Science and Discovery lab. Read the Science and Discovery lab brief first.
2. Ask the practitioner: "Which part of your work would you never hand to AI, and why?" Note their answer and their role in your Future Map.

About the evidence

ILO scores (a global generative-AI task-exposure estimate; not a Malaysian forecast; not a job-security rating), on a 0 to 1 scale with the ILO's band^[1]:

• Generalist medical practitioners 0.29, not exposed
• Nursing professionals 0.25, not exposed
• Health care assistants 0.14, not exposed
• Medical imaging and therapeutic equipment technicians 0.22, not exposed
• Physiotherapists 0.28, minimal exposure
• Medical and pathology laboratory technicians 0.31, minimal exposure
• Pharmacists 0.33, minimal exposure
• Medical records and health information technicians 0.52, gradient 3
• Medical secretaries 0.53, gradient 3
• Mathematicians, actuaries and statisticians 0.56, gradient 3

The bands run from "not exposed" and "minimal exposure" up to gradients 1 to 4, the most exposed^[1]. The ILO built its scores from a survey of 1,640 workers in Poland, rounds of discussion with international experts, and an AI assistant built on that input. Every task counts equally, and the ILO says local factors its global study does not cover can speed up or limit the change^[1]. ILO occupation groups do not map one-to-one onto Malaysian job titles, and we do not claim a direct mapping.

WEF figures are expectations from a survey of more than 1,000 employers across 55 economies, published January 2025^[2]. Anthropic figures are its own usage data or a US model, not Malaysian demand. The method behind the task cards is in Tasks, not job titles.

For facilitators

Use the task cards as a starting list for the Day 2 AI task X-ray and ask the practitioner to correct them before learners use them. Each learner supports one card with a source and challenges one assumption. Stop any conclusion that a job is "safe" or "dead": the evidence here does not support either. A clinician should check the X-ray card wording before cohort use.