Malaysia AI / AI news for Malaysia
UTM sets target for 30 per cent of work tasks to use agentic AI
UTM's Vice Chancellor says 30 per cent of work tasks should run on agentic AI, with responsible officers still reviewing the output. The column names no start date, tools or baseline, so the target is hard to measure yet.
In brief
- On 11 October 2026 UTM Vice Chancellor Professor Ts. Dr. Mohd Shafry Mohd Rahim published a column saying UTM has set a target for 30 per cent of work tasks to use agentic AI.[1]
- Examples include meeting minutes, periodic reports, official letters, document filing and basic enquiries. AI output must still be checked and approved by a responsible officer.[1]
- UTM Digital is now the Office of Service Quality and Digital Transformation, and every responsibility centre is asked to pick at least one process for agentic AI.[1]
What UTM announced
Universiti Teknologi Malaysia has set a target for 30 per cent of its work tasks to be done with agentic AI. Vice Chancellor Professor Ts. Dr. Mohd Shafry Mohd Rahim set it out in a column on UTM NewsHub, published on 11 October 2026 under the Professor's Insight label.[1]
The column is a leadership statement, not a policy document or a progress report. It reports no results, and this article does not claim that any task has yet moved to AI at UTM.[1]
What the 30 per cent target covers
The column defines agentic AI as technology that carries out a task from instructions and set steps, not just answers questions. It suits repetitive work with clear rules. Its test is simple: if a task can be explained step by step to a new employee, part of it could be done by agentic AI.[1]
The Vice Chancellor lists nine kinds of routine work. They include meeting minutes, periodic reports, official letters, filing documents by classification, and answering basic enquiries from students, staff and the public. He expects a lighter workload, higher productivity and better service quality.[1]
The column does not say how the 30 per cent will be counted. It does not state whether the share is measured by hours, by number of tasks or by process, and it names no baseline or deadline.[1]
Human review stays in the loop
The column says AI-generated work still needs review and confirmation by the responsible officer, to keep information accurate. It repeats later that reviewing, judging and approving stay with people.[1]
In its own words, “AI bukan diperkenalkan untuk menggantikan manusia, sebaliknya meningkatkan nilai sumbangan mereka.” In English, AI is not introduced to replace people but to raise the value of their contribution. Student welfare, staff wellbeing and major organisational decisions, it adds, still need judgement and empathy.[1]
The column also commits UTM to staff training so that no one is left behind, and to reviewing work procedures with data security and confidentiality kept in place.[1]
The Smart Counter and the office behind it
One example is the Smart Counter, or Kaunter Pintar. Basic questions on student applications, claim procedures and university services could be answered at any time, using information from official university sources. Complex matters would still go to an officer by appointment. The column gives no launch date.[1]
UTM has changed UTM Digital into the Office of Service Quality and Digital Transformation, which the column says focuses on service quality and not only on systems. Its Digital Innovation Centre is building applications for academic management, administration and services, but the column names none.[1]
Every faculty, department and responsibility centre, known as a PTJ, is asked to find repetitive tasks and choose at least one process for agentic AI. The column links the effort to the Malaysia Higher Education Plan 2026 to 2035. It says success will be judged by how far technology eases student matters and staff workload.[1]
Why Malaysia should care
A Malaysian technology university has set a numeric AI target for its own back-office work. Other organisations run similar paperwork, so the review rule and the task-by-task approach are the parts that transfer.
UTM staff and students
Routine paperwork and basic enquiries are the first targets, and complex cases stay with officers.[1]
Practical move: Watch for faculty announcements on which processes move first.
Other Malaysian universities and public bodies
A named percentage gives a clear goal, but without a baseline it is hard to compare.[1]
Practical move: Define how the share is counted before announcing one.
SME owners with admin-heavy teams
UTM's test, can you explain the task step by step to a new hire, works in any office.[1]
Practical move: Pick one repetitive task and name who reviews the AI output.
What Malaysians can do now
- Managers: list five repetitive tasks you could explain step by step to a new employee, then test one with AI.
- Team leads: name the person who reviews and approves each AI-generated output before it reaches customers or management.
- Compliance officers: set out what company data staff may enter into AI tools before agents handle routine work.
What we still do not know
What we still do not know
- How UTM will measure the 30 per cent, and by what date, since the column gives no baseline or deadline.
- Which agentic AI tools UTM will use and where its data will be stored.
- Whether and when the Smart Counter will launch, since the column describes it as an example.
Sources
- 1.Bekerja lebih bijak bersama AI, mentransformasi budaya kerja universiti UTM NewsHub, Pejabat Naib Canselor, 11 October 2026
- 2.File:Sultan Ibrahim Chancellor Building.jpg Wikimedia Commons, 2 April 2018
- 3.File:Sultan Iskandar Hall.jpg Wikimedia Commons, 2 April 2018
- 4.File:Universiti Teknologi Malaysia (Entrance).jpg Wikimedia Commons, 4 October 2026
More AI News
View all
