SINGAPORE/MALAYSIA: Synthetic intelligence (AI) now isn’t simply writing essays or making paintings anymore. It has developed into fronting convincing, human-like social media personalities that may communicate, smile, and acquire belief whereas spreading faux data.
A Channel NewsAsia (CNA) investigation discovered round 500 TikTok movies that includes AI-generated or manipulated girls making false or deceptive claims about Singapore and Malaysia. Printed between October 2025 and June 2026, the movies attracted greater than three million views earlier than some accounts have been eliminated, CNA reported (July 13).
The movies lined topics corresponding to Singapore’s economic system, overseas coverage and ties with China. Whereas they combined actual occasions with real statistics, many slipped in claims that have been both distorted or fully fabricated, making the misinformation tougher to identify.
Acquainted faces delivering the identical script to affect/persuade viewers
CNA examined 30 TikTok accounts and located a production-line method behind greater than 550 movies. Virtually all used AI-created, manipulated, or copied feminine presenters. Many additionally recycled voices, scripts and speaking factors throughout completely different accounts.
One false story claimed that Singapore Overseas Minister Dr Vivian Balakrishnan had unsuccessfully pleaded with China and Indonesia to cease a delivery route that might bypass Singapore’s port. CNA discovered no proof that such an incident ever occurred.
Of the 550 movies analysed, 94 repeatedly promoted the identical deceptive claims. Collectively, these movies alone drew greater than 1.6 million views.
The investigation discovered that 24 completely different accounts rotated via an identical scripts over a number of weeks, creating the impression that a number of unbiased creators had reached the identical conclusion.
After CNA shared examples with TikTok and sought remark, the platform eliminated two of the accounts for violating its guidelines on misleading behaviour.
TikTok stated it actively removes accounts concerned in actions corresponding to covert affect operations, impersonation, spam and different types of manipulation.
Actual information used to assist false tales
Somewhat than inventing the whole lot from scratch, many movies started with real headlines earlier than steering viewers in the direction of deceptive conclusions.
Some instructed Singapore’s port was about to lose its world place as a result of China’s Hainan Free Commerce Port had opened. Others pointed to ExxonMobil’s deliberate refinery closure as supposed proof that Singapore’s economic system was in decline.
These claims ignored essential info. Singapore dealt with a file 44.66 million containers in 2025 and has remained the world’s second-busiest container port for the previous 15 years.
The identical technique appeared in movies discussing funding hyperlinks with China. Whereas Singapore has been China’s largest overseas investor since 2013, some creators used that truth to counsel Singapore’s success depended solely on aligning with Beijing.
Malaysia was additionally focused. A number of movies blamed the nation’s slower improvement on insurance policies affecting its ethnic Chinese language neighborhood.
Specialists warn that AI-generated influencers use repetition to construct viewers belief
Affiliate Professor Saifuddin Ahmed from Nanyang Technological College’s Wee Kim Wee College of Communication and Info stated these campaigns don’t have to steer folks instantly.
As an alternative, repeating the identical message throughout many accounts makes the claims really feel extra acquainted, rising the possibility that viewers settle for them over time.
The Affiliate Professor additionally stated engaging AI-generated presenters assist maintain viewers watching, utilizing lots of the identical psychological strategies that make influencer content material profitable.
Benjamin Ang, who heads the Centre of Excellence for Nationwide Safety on the S. Rajaratnam College of Worldwide Research (RSIS), stated Mandarin-speaking audiences might face additional challenges as a result of fact-checking efforts in that language nonetheless path behind English-language content material.
Mr Ang inspired folks to assist family and friends recognise the risks of countless doomscrolling, as an alternative of feeding them deceptive content material to spice up views and engagement.
Mr Saifuddin added that making an attempt to identify AI-generated faces by itself is turning into unrealistic because the expertise improves. A greater defence is studying how coordinated campaigns work, corresponding to recognising repeated claims that seem throughout many accounts however aren’t reported by credible information organisations.
AI makes coordinated disinformation cheaper and tougher to detect
The investigation means that trendy affect campaigns not depend upon a single viral submit. They will depend on dozens of convincing AI personalities repeating the identical narrative till it feels acquainted and even convincing.
Whereas CNA couldn’t establish who operated or funded the community, the sample confirmed how cheaply AI can produce polished content material at scale, making misinformation tougher to detect and simpler to unfold.
Singapore has invested closely in digital literacy over time, however expertise retains altering. Studying to query repeated narratives, examine trusted sources and pause earlier than sharing content material often is the easiest defence any of us has. As AI turns into extra convincing, wholesome scepticism is turning into simply as helpful as digital abilities.
Learn associated: Police uncover footage of AI-generated faux Zoom assembly rip-off impersonating PM Wong that tricked sufferer out of S$4.9M
Learn extra: ‘Social cohesion takes a very long time to construct; it may be damaged if not cautious’ — Josephine Teo urges Singaporeans to not re-share divisive social media content material
