Malaysia's platforms were asked to take down 174,099 scam posts this year, up from 102,113 all of last year

AI Summary

Malaysian social media platforms were asked to remove 174,099 scam-related posts as of 15 October 2026, against 102,113 for the whole of the previous year, according to MCMC statistics cited by Deputy Communications Minister Teo Nie Ching. Platforms removed 170,633 of this year’s posts, a compliance rate of about 98 per cent. The same statement warns that AI-generated distress images are now used to solicit donations. The practical defence is to verify the appeal and the bank account before giving.

What happened

Social media platforms in Malaysia received 174,099 requests to remove scam-related posts as of 15 October 2026. Throughout the previous year, the figure was 102,113. The numbers come from Malaysian Communications and Multimedia Commission (MCMC) statistics, cited by Deputy Communications Minister Teo Nie Ching and reported by Malay Mail on 2 November 2026.

Of this year’s requests, 170,633 posts were removed. That is a compliance rate of about 98 per cent, meaning platforms acted on almost every request the regulator sent them.

The same statement carried a warning about a specific pattern: AI-generated sympathy scams, in which fake images of people in distress are used to solicit donations. The report does not attach a count to that pattern. It names it as a method, not as a measured share of the total.

Takedown request: a formal message from a regulator to a platform asking it to remove a piece of content. It is not a court order. The platform decides whether to comply, and the compliance rate measures how often it does.

The MCMC is the regulator that handles communications and multimedia in Malaysia, including complaints about online content. A takedown request is the tool it uses against a post it has assessed as a scam.

Why now

The thread is two to three years long. Scam content on social platforms has moved from crude text offers to material that looks like a real person asking for help, and the regulator’s response has scaled with it. Last year’s 102,113 requests were already a large number. This year passed that total before the year was three-quarters done.

What surfaced this month is the pairing. The takedown figures and the AI-image warning arrived in the same statement, which tells a reader that the two are connected in the regulator’s thinking. Cheap image generation has lowered the cost of producing a convincing distress photograph. A scammer no longer needs a real victim’s picture; a generated one does the same work and carries no trail back to a real person.

The second reason it surfaced now is arithmetic. A year-on-year jump of this size, visible by mid-October, is a signal that the annual total will land well above last year’s. A regulator that publishes the running figure is telling platforms and the public that the volume is not settling.

So what

The gain lands on the platforms, which now have a documented compliance rate of about 98 per cent. That number is a defence: it says the companies act when asked. The cost lands on the regulator and on the public, because a takedown request is a response after a post is already live. Someone has to see the scam, report it, have it assessed, and wait for removal. The person who pays in that gap is the reader who donates before the request is sent.

The mechanism is a loop with a delay. A scam post goes up, it reaches an audience, some fraction of that audience sends money, the post is reported, a request goes out, the platform removes it. The delay sits between the post going live and the removal, and the money moves inside that delay. A 98 per cent compliance rate says the loop closes. It says nothing about how fast, and speed is what decides whether a household loses money.

The change in the world is the input to that loop. AI-generated images make the appeal cheaper to produce and harder to dismiss on sight. A reader who has learned to distrust an obvious stock photo has not learned to distrust a generated one. The defence moves from judging the picture to checking the account.

What it means for Malaysia

The consequence lands on the household that gives, and on the small business or individual whose face is used without consent.

For the household, the carrier is the bank account behind the appeal. A genuine donation drive routes money to an organisation with a name that can be checked. A scam routes it to a personal account, often opened recently, often in a name that has no connection to the cause. The step a reader can take is to ask for the account name and compare it with the organisation’s registered name before transferring anything. That check costs a minute and does not depend on any regulator.

For the person whose photograph is used, the carrier is a police report. A generated image built from a real person’s face puts that person’s likeness behind a fraud they did not commit. The report in the source names the pattern; the practical step for a reader who finds their own photo attached to an appeal they never made is to lodge a police report so the misuse is on record.

For the platform user, the carrier is the report button. The 174,099 requests did not originate with the regulator alone; they follow reports and assessments. A reader who reports a suspicious post is feeding the loop that produced this year’s figure.

The wider Malaysian context is that digital fraud has become a routine cost of being online. The same week’s reporting on school funding in Kulai shows the other side of the same connectivity: classrooms getting interactive smartboards and a new RM4.3 million block at SJKT Kulai Besar, which serves 830 pupils. The infrastructure that carries a lesson also carries a scam. Readers following the technology thread can see the same pattern in how Malaysia’s digital economy is being built out, from the Penang technology finance centre timeline to the record RM215 billion export month that shows how much of Malaysian life now runs through online channels.

How confident

Confirmed. The MCMC statistics, cited by Deputy Communications Minister Teo Nie Ching and reported by Malay Mail on 2 November 2026, state 174,099 requests to remove scam-related posts as of 15 October 2026, against 102,113 for the previous full year. They state that 170,633 posts were removed this year, a compliance rate of about 98 per cent. The same report records the warning about AI-generated distress images used to solicit donations.

Inferred. The jump in requests points to a rise in scam content reaching Malaysian audiences, and the pairing of the figures with the AI-image warning suggests the regulator sees generated images as part of that rise. The mechanism is that cheaper image production lowers the cost of a convincing appeal, which raises the volume of posts worth reporting. This is an inference from the two facts sitting in one statement, not a measured link. The source gives no count for AI-image scams and no time-to-removal figure, so the size of that share and the length of the exposure window are unknown.

What would change my mind. A published breakdown of the 174,099 requests by scam type, with AI-generated imagery as its own line, would confirm or shrink the inference. A median time-to-removal figure would tell a reader how long the exposure window actually is. And a figure for money lost to donation fraud, separate from the takedown count, would show whether the loop is closing faster than the losses are growing. None of those three numbers appears in the source.

Also read: VMware’s new vSphere Express gives small Malaysian IT firms a cheaper path to stay on VMware

Also read: Sony revives 1979 Walkman headphones with USB-C — nostalgia meets modern audio in Malaysia

Also read: Nvidia CEO scandal exposes chip flows through Singapore-linked firms — what it means for Malaysia’s semiconductor plants

For you

• A household asked to donate to a distress appeal: ask for the account name and match it against the organisation’s registered name before transferring; a personal account behind a public cause is the warning sign.
• A parent whose teenager follows charity accounts on social media: agree a rule that any donation goes through an organisation the family can name, not through a direct transfer to an individual.
• A small business owner whose staff handle the company’s social accounts: treat a request to share or boost a donation appeal as a decision that needs a check on the account behind it, because a shared scam reaches your customers under your name.
• Anyone who finds their own photograph attached to an appeal they never made: lodge a police report so the misuse of your likeness is on record, and report the post to the platform.

The data point

174,099 — requests to remove scam-related posts sent to social media platforms as of 15 October 2026, against 102,113 for the whole of the previous year.

How we calculated this

The source gives two request totals and one removal total. The arithmetic below uses only those figures.

Requests this year (as of 15 October 2026):   174,099
Requests previous full year:                  102,113

Increase in requests:  174,099 - 102,113 = 71,986
Increase as a share of last year's total:  71,986 / 102,113 = 0.705 = about 70.5 per cent

Posts removed this year:                      170,633
Requests not removed this year:  174,099 - 170,633 = 3,466
Removal share:  170,633 / 174,099 = 0.980 = about 98 per cent

Read plainly: this year’s requests passed last year’s full total by 71,986, or about 70.5 per cent, with the year still running at the point of measurement. The 3,466 requests without a recorded removal are the gap the 98 per cent figure describes.

FAQ

What do the 174,099 requests actually measure?

They measure requests sent by the Malaysian Communications and Multimedia Commission to social media platforms asking for scam-related posts to be removed, as of 15 October 2026. They do not measure the number of scams in circulation, the number of people who saw a scam, or the amount of money lost. A single scam campaign can generate many posts and therefore many requests, so the figure counts posts, not criminals and not victims. Read it as a volume indicator for enforcement activity, not as a victim count.

Why does the compliance rate matter to an ordinary reader?

A compliance rate of about 98 per cent means platforms removed 170,633 of the 174,099 posts the regulator asked them to remove this year. For a reader, that says the takedown system works once a post is identified. It does not say how quickly removal happens, and the source gives no time-to-removal figure. The money in a donation scam moves in the gap between a post going live and its removal, so speed matters more to a household than the final compliance number.

What is an AI-generated sympathy scam?

It is a donation fraud that uses images produced by artificial intelligence to show a person or an animal in distress, paired with an appeal for money. The image is not a photograph of a real victim. Because generating an image is cheap, a scammer can produce many appeals and vary them. The defence is not to judge the picture but to check the destination: ask for the account name, compare it with the organisation’s registered name, and give only through a channel you can verify.

The point

The takedown numbers show a system that works after the fact: 174,099 requests, 170,633 removals, about 98 per cent compliance. What they do not show is the delay, and the delay is where a household’s money goes. AI-generated distress images shorten the time it takes to produce a convincing appeal and lengthen the odds that a reader gives before anyone reports the post. The check that survives all of that is the account name, and it costs a minute.