How to Grow Malaysian Workers' Confidence in AI

Malaysians are adopting AI faster than they are being trained to use it. Here is what the data says, what builds confidence, and what we learned about it in Nigeria.

The briefing: adoption is up, confidence in using it is down

ManpowerGroup's Global Talent Barometer 2026 found that 60% of Malaysian workers now use AI tools regularly, seven percentage points more than the year before. Confidence in technology fell seven points over the same period, to 78%. The survey covered 536 Malaysian workers, interviewed between 1 September and 1 October 2025. Overall sentiment improved, with the Malaysian barometer score at 73%, up three points. The problem is specific to technology confidence (Malay Mail).

The drop was uneven. ManpowerGroup's Malaysia report shows Gen X confidence in technology down 21 percentage points while millennials rose 2 points. Globally, the same pattern holds: Baby Boomers' tech confidence fell 35%, and Gen X's fell 25% (ManpowerGroup). The people who most need to feel capable are losing confidence fastest.

Why confidence is falling

Training has not kept up. 55% of Malaysian workers reported no recent training, and 32% had no access to mentorship. EY's 2025 Work Reimagined Survey found the same gap from a different angle: only 12% of Malaysian employees receive sufficient AI training to fully benefit from AI, even though 81% of active GenAI users report significant time savings. Regionally, Milieu Insight's February 2026 survey of 3,000 workers in six Southeast Asian markets found employer-provided training reached only 14% to 29% of workers, depending on the country.

Older workers are being skipped. Randstad's 2025 Workmonitor found 41% of Gen X and 58% of Baby Boomers said their employers introduced no new training in the past year. Randstad also reported 35% of all Malaysian respondents said their employers had introduced no new training in the last 12 months, while 41% were interested in AI training if offered, up 12% from 2024 (New Straits Times, The Iskandarian). The demand is there.

Anxiety and workload are rising. 48% of Malaysian workers feared automation could make their roles obsolete within two years, and 30% expected possible job loss soon. EY found 68% say their workloads have grown. People under that pressure rarely have time to learn something new on their own.

What builds confidence: five moves

Start from the worker's own problem, not the tool. Ask each team what slows their day down, then show how AI addresses that task. Confidence grows when someone solves a real problem in a real afternoon.
Train by role and by level. An executive, a team lead and a frontline operator need different things. A single generic session leaves each group feeling it was meant for someone else.
Build mentorship in. With 32% of workers reporting no access to mentorship, pairing confident users with hesitant ones is cheap. It is especially useful for Gen X, whose confidence fell furthest.
Say what is allowed. Governance and security worry leaders as much as workers. Clear rules on data, approvals and what AI may not do remove a major source of hesitation.
Measure confidence, not attendance. Track whether people use the tool, whether their output is correct, and whether they say they would use it again. Course completion rates do not show that.

What Malaysia is already funding

The public support is considerable, and many employers have not used it yet:

The Ministry of Digital's Budget 2027 plan aims to produce 200,000 AI-skilled individuals through AI Malaysia Bhd, to help 4,000 MSMEs adopt AI through MDEC funds, and to certify 5,000 AI professionals (Malay Mail).
TalentCorp, with Ernst & Young, is offering 22,000 free AI training places nationwide through the Jelajah AI Mahir roadshow (Bernama).
A Budget 2027 commentary reported that under the current 2026 measure, eligible MSMEs can claim a further 50% tax deduction on recognized AI training expenses (The Sun). Confirm eligibility with your tax adviser.

The stakes are real. Malaysia's Human Resources Minister warned that nearly 700,000 jobs could be significantly disrupted by AI, digitalization and the green economy within five years unless workers upgrade their skills (People Matters).

What we learned in Nigeria, and how Agents Stack can help

Malaysia is not Nigeria, and the two markets differ in many ways. But the confidence problem looked familiar to us when we ran an executive AI training there.

We designed that training so every level of a company could follow it, with use cases for each level. The core exercise was to identify the biggest problem in each leader's daily work and test whether AI could solve it. Executives were keen to adopt AI because they did not want to fall behind. Their main concern was governance and security.

Our biggest takeaway was that tools were not the bottleneck. What mattered was whether people would adopt them, and whether the company had a team that could drive that adoption. We also saw how much a small, honest room can accomplish: ten people came to a community event, Brew the Future, and the conversation was strong enough that we expect to work together. These are our observations, not survey results.

For Malaysian employers, Agents Stack proposes to apply the same approach:

Role-based AI training for executives, managers and teams, built around each group's real workflows.
Use-case workshops that start with the problems staff name, not a tool list.
Governance and security guardrails written into the training, so people know what they may do with company data.
Advisory and instrumentation to measure adoption and confidence over time, so leaders can see who is stuck.
Mentorship pairing to bring Gen X and other less confident staff along with their peers.

This is a proposed program. We are not claiming verified results from Malaysian customers.

Monday's agenda

Ask five staff, including at least one Gen X employee, what their hardest task is and whether AI could help.
Check which of the public schemes above your company qualifies for.
Write one page on what staff may and may not do with AI and company data.
Pick one team and one workflow for a 30-day pilot, with a named mentor.
Measure usage, accuracy and self-reported confidence at the start and at day 30.

What one task would your team most like AI to take off their plate?

Book 30 minutes →

Sources

Written by

Precious Udoeka Godfrey

Contributor, ASEAN

Precious Udoeka Godfrey writes about AI adoption and workforce readiness across ASEAN for Agents Stack.

Connect on LinkedIn →