Every few weeks, another large telecom operator announces that it is shrinking, and the reason given is increasingly the same: automation, now accelerated by artificial intelligence. The cuts have mostly landed in Europe and the United States so far. For Malaysians who work in telecoms, or in any desk job built on routine tasks, the pattern is worth understanding, because a version of it has already arrived here.
Editor
Kai T chevron_right
Table of Contents
The cuts are real, and they run deep
Across 20 large operators it has tracked for more than a decade, Light Reading reported on 16 June that almost 160,000 jobs have disappeared since generative AI arrived in late 2022, a drop of about 11 percent in roughly three years. More than 63,000 of those roles went last year alone. Verizon chief executive Dan Schulman, who removed 10,000 jobs in his first quarter in charge, told the Wall Street Journal he expects unemployment to climb as high as 30 percent over the next two to five years.
The picture elsewhere is similar. Britain's BT plans to employ no more than 80,000 people by 2030, down from over 107,000 in March, while Norway's Telenor has fallen from around 37,000 staff a decade ago to roughly 10,000 at the end of 2025. Not one of Europe's big telcos has grown its headcount over that period.
There is an important caveat, and the same report is careful to make it. Most of this shrinkage so far owes less to AI than to operators selling off divisions, outsourcing work, and switching off older networks. Labour is only about a fifth of a big telco's running costs, so swapping staff for software does little for profit on its own. The worry is that the pull of AI is becoming hard to resist regardless of the arithmetic.
What it means for Malaysia
Malaysia has already lived through its own telco downsizing. After Celcom and Digi combined in late 2022 to form CelcomDigi, the country's largest mobile operator, it ran a voluntary separation scheme that cost about RM140 million in the first quarter of 2024, part of an integration targeting RM8 billion in net synergies. The company says network and IT integration was more than 90 percent complete by September 2025 and expects RM700 million to RM800 million in annual savings after 2027. Leaner networks tend to need fewer hands to run them.
The pressure reaches well beyond telcos. A TalentCorp study released in 2025 found that about 620,000 Malaysian jobs across 10 major sectors, including information and communications, sit at high risk of disruption from AI, digitalisation, and the green economy. Another 1.8 million were rated medium risk. The same study flagged roughly 60 emerging roles, most of them in AI and digital fields, and the government has since rolled out reskilling drives such as the Jelajah AI MyMahir programme, run with EY Malaysia from January 2026.
A bet that cuts both ways
Malaysia's wider ambition sharpens the irony. The country is courting data centres and casting itself as an AI hub, the same build-out we weighed when we asked whether Malaysia can afford its AI bill. Yet the automation that justifies those investments is also what thins out operational jobs. There is a sovereignty wrinkle too. Most of the leading AI models are controlled abroad, and Light Reading notes a recent move by Washington to restrict some non-citizens' access to a top American AI system, a reminder that leaning on foreign models carries its own risk for any country, Malaysia included.
For Malaysian tech workers, the takeaway is preparation rather than panic. The jobs that last will sit closest to the AI doing the work, and the gap between those who reskill and those who do not is only going to widen.
Images courtesy of David Arrowsmith and Arlington Research on Unsplash.