AI Governance & Regulation · 2026-07-17 · 02:17

The AI 'news anchors' spreading misleading claims about Singapore on TikTok

Speaker
CNA
CNA report
Type
Industry Leader
Source
CNA

In Brief

CNA's investigation uncovered over 550 TikTok videos using AI-generated female personas, with nearly 9 in 10 spreading false or misleading claims about Singapore and Malaysia.

Readable transcript

Caption language: en · Fetched: 2026-07-17

They look like presenters, they sound authoritative, and they speak confidently about Singapore and Malaysia. But here's the catch. Many of them don't actually exist. Over the past 9 months, a CNA investigation has uncovered more than 550 TikTok videos built using AI-generated, manipulated, or copied female personas. Almost nine in 10 of them push false and misleading claims. Among the more outrageous claims, one video falsely claims Singapore's Foreign Minister Vivian Balakrishnan had backed China and Indonesia not to let a new shipping route bypass Singapore's port. It never happened. The thing is, these videos don't look like propaganda. They look more like ordinary geopolitical or financial commentary, and that makes them harder to spot. Together, the videos racked up more than 3 million views.

Many start with real news before distorting facts and ending in a misleading conclusion. [music] One claims Singapore's port was on the verge of collapse because of a new Chinese shipping hub. Another used ExxonMobil's plant closure as supposed proof that Singapore is in decline. Neither claims reflect reality. CNA also found the same scripts being repeated by different presenters. Sometimes different accounts use the exact same voice. Other times, the same face appeared with different voices. The result? It looked like independent creators reaching the same conclusions when they instead appeared to be part of the same content production line. Experts say that's exactly the point. The campaign doesn't have to convince people all right. It just has to make certain narratives feel familiar.

Experts warn repeated narratives like these can erode trust and deepen divisions over time. Experts also say spotting fake faces or AI glitches is becoming harder. Instead, you should watch for the playbook. Different accounts repeating the same talking points, claims that aren't reported anywhere else, and repetition replacing evidence. Because while the faces may change, the tactics often don't.

Related Videos

Singapore introduces generative AI guidelines on use of personal data, chatbots

2026-07-20 · Aldrina Thirunagaran · 10:40

Singapore introduces new guidelines for generative AI use of personal data, aiming to help businesses adopt AI responsibly while building public trust.

Singapore Streamlines Construction Robotics Regulations

2026-07-17 · Nikhil Khattar · 03:02

Singapore streamlines construction regulations to hasten robotics adoption, supporting firms to scale successful trials and identify trusted technologies.

Tourist guides adapt as AI and social media reshape travel habits

2026-07-11 · CNA · 03:35

As travelers increasingly rely on AI-generated itineraries and social media recommendations, tourist guides face fewer opportunities and must innovate to adapt.

Man in the loop a must when using AI in defence

2026-06-26 · Chan Chun Sing · 01:12

Singapore's Defence Minister Chan Chun Sing argues that humans must remain in control of decision-making when AI is used in military operations.

Masagos: No Vulnerable Group Should Be Left Behind in AI Push

2026-05-08 · Masagos Zulkifli · 02:33

Minister for Social and Family Development Masagos Zulkifli, speaking at the St Gallen Symposium, stressed that AI rollouts must include vulnerable groups so that the technology gains do not bypass them.

Some training providers see enrolments in AI courses double

2026-05-02 · CNA · 02:55

Training providers report sharp rises in AI course enrolments, with some seeing numbers double. Stronger employer support is needed to translate AI skills into workplace capabilities.

More on these topics