MALAYSIA stands at the crossroads of technological change. Generative artificial intelligence (AI) is rapidly transforming jobs, from clerical roles to professional services. While new opportunities will emerge, there is also the risk of displacement, particularly in routine and repetitive tasks.
The recent study by the Institute of Strategic and Inter-national Studies, the World Bank and think tank Konrad-Adenauer-Stiftung Malaysia highlights complementary policy directions for Malaysia: strengthening social protection, reforming education and lifelong learning, and realigning labour market institutions.
Together, these measures aim to safeguard workers while preparing them for the future of work.
> A safety net for all workers
The study warns that Malay-sia’s current safety nets leave many workers exposed. Employees in the formal sector benefit from schemes under the Employees Provident Fund (EPF) and the Social Security Organi-sation (Socso). Yet, millions of self-employed and informally employed workers remain unprotected.
To close this gap, the study recommends extending unemployment insurance to all workers, not just those in traditional employment. A voluntary scheme, modelled on EPF’s i-Saraan with government co-contributions, could encourage informal workers to opt in.
Even so, experts caution that social insurance cannot reach everyone. Here, non-contributory social safety nets – funded by public revenue rather than payroll deductions – are vital.
Malaysia already runs extensive cash transfer schemes, such as Sumbangan Tunai Rahmah covering about 66% of households. But the study notes that benefit levels remain modest, averaging just 9.5% of pre-transfer income for the bottom 20% of households. This falls far short of the 19% average in high-income economies.
Adding to the challenge is fragmentation: more than 155 federal programmes are delivered by 18 agencies, each with different rules. Consolidating this patchwork system into a more accessible “one stop” front-end would make support easier to navigate for struggling households.
We need more complementary policies aimed at safeguarding workers while preparing them for the future of work. —123rf
> Lifelong learning as a shield
If AI is to be harnessed for growth, Malaysians must be ready to work with it rather than against it. The report stresses the need to streamline lifelong learning opportunities. Currently, multiple agencies – from the Human Resources Development Corp (HRD Corp) to the Malaysia Digital Economy Corporation – offer overlapping training schemes. While choice is abundant, duplication and inefficiencies make the system harder to use.
A single digital platform could guide workers towards the most relevant courses based on their skills and career aspirations. National recruitment agency TalentCorp’s MyMahir.my is seen as a promising start, but the study urges continuous user feedback, better accessibility, and outcome tracking to ensure training remains relevant.
In the longer term, the funding model must evolve. Today, training is largely tied to employer levies under HRD Corp. Reform could include individualised learning credits through a portable “skills wallet”, allowing workers to accumulate training rights across jobs and careers – an approach already in practice in Singapore.
> Rethinking education
Preparing the next generation requires embedding AI into the classroom without undermining critical thinking. The report calls for mainstreaming AI education in schools, much like ICT was introduced in the 1990s. But it warns that this should complement – not replace – basic numeracy, literacy, and problem-solving skills.
Teachers, too, need upskilling. Expanding the AI Education for Educators programme into a compulsory, stackable certification track would ensure teachers stay ahead of the curve. Mean-while, bridging the digital divide by guaranteeing minimum connectivity and devices in rural schools is essential to avoid uneven opportunities.
At the same time, the study urges schools to double down on “human-edge” skills such as creativity, social intelligence, and emotional reasoning – areas where humans still outperform machines. Project-based learning, cross-school collaborations, and stronger language arts curricula could strengthen these capacities while shaping more adaptable, civic-minded graduates.
> Incentivising employers
Finally, the study highlights the need to realign labour market incentives. At present, payroll- linked social protection makes hiring workers more expensive than investing in machines. Adjusting tax structures or introducing targeted credits could encourage companies to retrain employees instead of replacing them with AI.
Job quality also matters. Occupations resistant to automation – like caregiving, trades, and personal services – remain undervalued. Raising wages, improving safety standards, and supporting new forms of worker representation would help these roles remain attractive and sustainable.
