Malaysia's labour market is beginning to demonstrate what the age of artificial intelligence (AI) looks like in practice: not a dramatic elimination of jobs, but a gradual reshaping of who gets employed, what talents are rewarded, and which sectors are advancing fastest. According to the most recent Malaysia cut of PwC's 2026 Global AI Jobs Barometer, the country is not experiencing a conventional AI hiring explosion. Instead, it sees something subtler, and possibly more significant. AI is becoming more prevalent in mainstream recruiting, even as overall job creation remains uneven and firms negotiate a softer post-2023 demand climate.

This distinction helps to explain the data. On the surface, the headline stats may appear lacklustre. The total number of Malaysian job listings requiring AI-related skills increased just slightly in 2025, to around 29,000 from 27,000 in 2024, remaining below the 33,000-peak reported in 2023. But beneath that level, another striking trend emerges: AI's percentage of all job postings has doubled, from 1.9% in 2024 to 4% in 2025. In other words, even if the actual number of people employed is not increasing, a greater proportion of employers now demand that employees are able to use, deploy, or collaborate with AI.

This shows that Malaysia's AI story is no longer limited to a few professional technical positions. The research implies that AI-related skills are becoming more popular in mainstream employment. Employers may not be posting many more opportunities overall, but they are requesting diverse talents inside those roles. That is indicative of diffusion rather than hype. It indicates a market in which AI is being integrated into existing workflows in finance, professional services, consumer enterprises, and industrial activities rather than being considered just as a technical function.

The broader labour market context helps explain why this move can appear contradictory. According to PwC data, job postings across all AI exposure quartiles peaked in 2023 before declining in 2024 and 2025. By 2025, the lowest AI-exposure quartile had fallen to roughly 0.87 postings for every posting in 2021, while the highest exposure quartile was at around 0.95. That relative closeness is significant: recent hiring weakness appears to have been influenced more by macroeconomic conditions than by AI exposure alone. Simply put, AI has not protected enterprises from cyclical caution. It has altered what employers expect from employees, but it has not halted the business cycle.

That said, the top quartile of AI-exposed occupations still has the highest absolute number of postings. In 2025, the category had over 413,000 job posts, much above the second quartile's 187,000, the third's 70,000, and the bottom quartile's 41,000. The fact that the most AI-exposed occupations continue to be the largest hiring pool, despite a significant reduction since 2024, tells its own narrative. A big portion of job postings are concentrated in occupations with more AI exposure, whether through analytics, digital tools, software-assisted workflows, or data-driven decision-making.

Sector patterns support this picture of unequal but growing adoption. Between 2021 and 2025, energy, utilities, and resources had the highest share of total job postings in Malaysia, accounting for 25.3%, followed by professional services (17.7%). Health and the public sector made substantially less contributions, with 2.1% and 0.3%, respectively. However, when the focus switches from overall hiring to AI intensity, technology, media, and telecommunications (TMT) emerges as the clear leader, accounting for the largest share of AI job listings inside its own sector. More importantly, all sectors saw an increase in AI job shares in 2025 compared to 2024.

That broad rise may indicate that AI in Malaysia is no longer a niche story. TMT may continue to lead because it is the most technologically intensive industry, but the spread across sectors shows that AI capabilities are increasingly being considered as a cross-functional business requirement. A bank seeking AI-literate employees, an energy company automating maintenance and forecasting, a professional services firm incorporating generative AI into research and drafting, and a consumer-facing company using AI tools in marketing and operations are all part of the same structural shift, even if their business models are very different.

Wages provide another indication of how employers value these qualities. PwC discovers positive AI salary premiums across major industries in Malaysia, with financial services outperforming at approximately 66%. Health is the next highest at 41%, followed by manufacturing and energy at 34%, professional services at 30%, TMT at 23%, consumer markets at 20%, and government and public sector at 31%. These are not minor distinctions. They contend that where AI-related abilities are scarce or strategically valuable, employers are willing to pay significantly more.

However, pay data advises against making overly simplified conclusions. Higher AI exposure may not guarantee the greatest premium. TMT, for example, has the highest AI hiring rate but not the highest wage premium. Health is only moderately detailed in the report, yet commands a high cost. This indicates that sector-specific shortages, legal constraints, domain expertise, and the practical difficulty of integrating AI into certain jobs all have an impact on wages. The premium is not only for understanding AI, but also for being able to apply it in commercially useful ways in the appropriate setting.

If there is one recurring theme across the paper, it is that AI is altering jobs from the inside out. Malaysia has a low but positive correlation of 0.16 between AI exposure and net skill change from 2021 to 2025, indicating that more exposed occupations see somewhat larger adjustments in skill requirements. The top quartile of AI-exposed occupations had the highest average net skill change, at 6.53, compared to 5.32 for the bottom quartile, 4.02 for the third, and 4.54 for the second. The relationship is not entirely linear, indicating that occupation-specific effects are still important. However, the overall trend is clear: the more exposed a job is to AI, the more likely it is that its skill mix will evolve.

That trend is even more noticeable when considering the number of "new" talents arriving in occupations. In 2025, the most AI-exposed quartile added an average of 90 new skills per occupation compared to 2021, whereas the second quartile added 53, the third added 28, and the worst quartile added 23. This is a clear indication that AI isn't simply replacing one technical talent with another. It is widening the set of skills that employers want, from data literacy and prompt-based tool usage to problem solving, judgement, workflow design, and communication.

This is where the Malaysian AI workforce story shifts away from coders and towards the average knowledge worker. According to the survey, AI user roles now dominate the industry. These are positions that necessitate AI literacy or applied AI skills rather than advanced model-building knowledge. In 2025, AI user roles increased by around 1,800, while AI developer roles decreased by 467, reaching their lowest point throughout the study period. Overall, AI user positions increased 6.9%, while AI developer roles decreased 23.6%. Across sectors, AI-related hiring remained focused on AI usage rather than development, with TMT having the greatest developer share and consumer markets having the largest user-role share.

This may have significant implications for lawmakers, educators, and companies alike. Malaysia's near-term AI workforce problem is not solely focused on generating large numbers of frontier AI engineers. It is about preparing a much larger workforce to use AI effectively in finance, operations, customer management, compliance, healthcare support, logistics, engineering, and business services. Developer expertise is still important, particularly if Malaysia wishes to develop local intellectual property and advance up the digital value chain. However, current labour-market evidence indicates that adoption skills, or the capacity to work successfully and ethically using AI tools, are in high demand.

This has ramifications for both the structure of opportunity and the risk of exclusion. If AI becomes a normal layer in white-collar and technical employment, people with limited access to training, digital exposure, or workplace support may face slower adaptation, even if their roles are not formally automated. One concern is that employees with greater access to AI training and support may benefit more quickly than those without similar access. Wage premiums, skill expansion, and the prevalence of AI user roles all trend in this direction.

None of this suggests that Malaysia should interpret the findings as alarming. The report itself does not demonstrate mass displacement. If anything, it suggests a labour market in transition rather than in collapse. However, it would also be incorrect to interpret the data as evidence that market forces alone are sufficient to manage the transition. The situation is more nuanced than that. Hiring volumes remain subject to broader economic weakness; AI demand is expanding, albeit unevenly; and the benefits of that spread appear to be greatest where skills, resources, and organisational preparedness are already concentrated.

Policy responses could range from moderate to purposeful. First, companies must stop seeing AI capabilities as the domain of their technology teams and instead incorporate organised AI upskilling into mainstream workforce planning. Second, universities, TVET institutions, and professional bodies should update curricula more frequently, not just for technical AI competencies, but also for related skills such as data interpretation, critical thinking, and process redesign. Third, authorities should focus on dissemination beyond the traditional digital winners. If sectors such as manufacturing, energy, and healthcare are already seeing wage premiums and increased AI intensity, then support for adoption cannot be limited to tech parks and software enterprises.

For workers, the message is less dramatic than many headlines suggest, but no less important. The safest assumption is that AI literacy is becoming a part of general employability, just as spreadsheet skills and digital communication once were. For businesses and governments, the guidance is similar: expanding access to AI literacy so that adoption remains broad and inclusive across the workforce. Malaysia still has time to develop AI as a productivity advantage rather than a source of uneven outcomes. The findings show that the window is open, but not indefinitely.

Dr Jack Ng Kok Wah is a senior lecturer at Multimedia University, Cyberjaya, and an HRDF-accredited trainer focusing on AI research in marketing, healthcare, and digital innovation. He serves as marketing management adviser to corporate at Supersafe Industries, where he provides strategic advisory support in marketing, branding, and digital innovation.