SHAH ALAM, Aug 30 — The use of artificial intelligence (AI) in fraudulent activities is expected to make scam content more prevalent and increasingly difficult to detect over the next few years.
Malaysian Cyber Consumer Association (MCCA) president Siraj Jalil said scammers now use agentic AI technology and autonomous systems to automatically generate fraudulent messages and content at scale.
The technology also enables scam content to be produced more realistically, making it harder for users to distinguish between genuine and fake content.
“Scammers also use techniques such as algo-speak to alter words, symbols, or content formats to evade detection by algorithmic systems and platform moderation.
“In the past, scam content was easier to identify. Now it is becoming increasingly complex because videos and other content are being produced more realistically, particularly with the emergence of AI,” he said when contacted by Media Selangor.
Siraj added that these developments pose a challenge to the authorities, particularly the Malaysian Communications and Multimedia Commission and platform providers, to detect and take down scam content more quickly.
Users also need to play their part by reporting suspicious content found on social media, rather than relying entirely on the authorities and platform providers to take action.
“If you see something suspicious, report it immediately. Do not leave it unchecked because a single interaction with scam content can result in more similar advertisements being targeted at you,” he said.
Siraj also advised users not to click links or interact with advertisements that seem too good to be true, especially those involving financial offers.
On August 22, Deputy Communications Minister Teo Nie Ching said an estimated 99,693 pieces of social media content containing fraudulent material, including scams, had been taken down as of July 15 this year.
The figure exceeded the 98,503 pieces of content taken down throughout 2025.
She said 106,268 requests to remove such content were also submitted to social media platforms, including Meta, TikTok, and YouTube, during the same period, compared with 102,113 requests throughout last year.
