Malaysia's Plan for AI in Every University Curriculum

By: Retave’s Technology Desk Editor

Listen to this article

Ask any dean how long it takes to get a new course through accreditation in Malaysia, and you’ll get a sigh before you get a number. That gap, between how fast AI moves and how slowly a curriculum is legally allowed to change, is exactly the bottleneck AI Nation 2030 is trying to break.

The question was never really whether Malaysian universities want to teach AI. It’s whether the system built to approve what they teach can keep up with it.

Steady progress, real constraints

The National AI Action Plan 2026-2030 (AI Nation 2030) is transparent about current progress, aiming to motivate stakeholders by outlining tangible steps, including funding and policy direction, thereby fostering confidence in the plan’s potential.

Courses take too long to get approved. Too many lecturers haven’t been trained to teach AI competently. Industry shows up inconsistently as a partner, and graduates often can’t see a clear line from what they studied to an actual job. Getting past that, the plan suggests, means dropping the scattered pilot-project approach for a single connected system that builds hands-on technical skills, responsible data handling, and AI ethics together, rather than just teaching students to write clever prompts.

The target

AI-Centric Higher Education, or E3, falls under the Ministry of Higher Education (Kementerian Pengajian Tinggi, MOHE). Distilled down, the pillar is about three things: more students genuinely engaging with AI coursework, lecturers noticeably better equipped to teach it, and a wider set of real placements that link graduates to AI-related employers. Four levers are meant to pull that off: reshaping what gets taught, investing in the people who teach it, sharing infrastructure so no campus foots the entire bill, and giving students genuine industry exposure. All of it is anchored in responsible, values-driven use of the technology.

Four moves to mainstream AI in the curriculum

What this looks like in practice

The finer print gets specific about sequencing. Every graduate, regardless of discipline, is supposed to leave with a working baseline understanding of AI, while a smaller group of specialists gets pushed further through deeper technical tracks. Short AI electives are meant to spread well beyond computer science departments, with institutions borrowing what’s already proven elsewhere rather than each experimenting from scratch.

Those incoming MQA benchmarks also pull double duty. They’re meant to speed up approvals for new AI courses and give existing ones a way to update continuously, so a syllabus doesn’t quietly go stale the moment the technology moves again.

Curriculum reform doesn’t stop at the lecture hall door

This pillar also ties classroom learning to actual industry work. One strand pushes universities toward joint projects with companies: capstone assignments built around live business problems, with teaching materials kept open so other campuses can reuse them rather than start over. A second strand focuses on structured internships with organisations where AI genuinely sits at the core of the business, coordinated through shared national platforms instead of each university’s own contacts. The goal is real exposure, not a token placement with an AI label slapped on it.

None of this holds together, though, if the people actually teaching the curriculum haven’t been trained to teach it. To address this, the plan includes targeted faculty development programs and certification initiatives to ensure high-quality instruction across institutions, which is a critical and challenging component of the reform process.