AI is changing engineering work in Malaysia. MOSTI said diploma skills still matter

Malaysia's Science Ministry used a university career talk to make a grounded point about AI and engineering: the technology may change the tools, but industry still needs people who understand the machines, processes and failures behind those tools.[1]

At UiTM's Pasir Gudang campus on 2 July 2026, Science Minister Chang Lih Kang said diploma-level engineering graduates remained important as technical workers who could operate, maintain and improve AI-based systems in real industrial environments. The statement was a direction for skills—not a guarantee about future job numbers or salaries.[1]

Operating an AI-enabled system means more than pressing start. A technician may need to check sensors, confirm that input data is sensible, watch production constraints and recognise when an automated recommendation conflicts with the physical process.[1]

Maintenance also expands. Alongside motors, networks and controls, teams may have to track data drift, model versions, calibration, access permissions and the conditions under which a system should fall back to manual control. Those responsibilities connect traditional engineering discipline with software and data literacy.[1]

Improvement is the third layer. Engineers who understand the process can identify where an AI model adds value, where it creates a new failure mode and which measurement should decide whether a change stays in production. That combination is harder to replace than shallow familiarity with one AI product.[1]

Chang said TVET curricula should be strengthened with robotics, AI and coding rather than remain limited to conventional technical skills. The phrasing matters: the newer subjects were additions to the engineering foundation, not replacements for mathematics, safety, materials, controls or hands-on competence.[2]

UiTM deputy vice-chancellor Juliana Johari added a wider graduate profile: students should master technology and AI while becoming innovators, value creators and ethical leaders. That sets a higher bar than tool use because it includes judgment about where automation belongs and who carries responsibility when it fails.[1]

For a course designer, the practical move is to put AI inside an engineering task. Students can diagnose a machine fault, compare a model's output with measurements, document uncertainty and design a safe escalation path. A standalone prompt-writing lesson cannot test those capabilities by itself.[1][2]

The BERNAMA report carried by Malay Mail said 12 ministries were involved in TVET implementation. Chang noted that TVET was not directly under MOSTI, although the ministry would support the agencies responsible for strengthening it.[2]

That division of responsibility means a career talk cannot change a programme on its own. New modules need a named owner, updated assessment, instructors, equipment, industry placements and evidence that graduates can perform the work employers require.[2]

It also makes employer participation important. Factories, utilities, infrastructure operators and engineering firms can define the failure cases and operating constraints that a classroom may otherwise miss. Their contribution should be specific enough to shape projects and assessments, not only provide a logo or guest speaker.[2]

MOSTI's report documented the speakers, themes and intended awareness outcome. It did not quantify how many engineering jobs would be created, changed or displaced by AI, which disciplines faced the largest shift, or what wage premium a new skill would command.[1][2]

Students should therefore treat the message as a preparation framework rather than a prediction. A useful portfolio would show an engineering problem, the data and model used, the safety and failure checks, the human decision point and the measured result. That evidence travels better than a certificate listing an AI tool.[1]

The strongest part of the Pasir Gudang message was its focus on real systems. Malaysia does not only need people who can ask AI for an answer. It needs engineers and technicians who can decide whether that answer belongs in a machine, a maintenance process or a production line—and keep the system safe when reality disagrees.[1]