LETTER | On a phone in Kota Bharu, Kelantan, a child is watching a grinning cartoon tiger hold up a bowl of cereal.

No one ever drew the tiger. The cereal looks nothing like what is in the box. Both were produced by artificial intelligence, tested against thousands of alternatives, and chosen because they were hardest for children to resist.

We tend to file this under food policy or marketing ethics. It belongs under children’s rights. A child is entitled to protection from misleading and exploitative content wherever it reaches them, from a supermarket shelf to a social media feed.

AI has quietly opened a gap in that protection.

A recent United Nations consultation on children’s digital rights identified the problem. Global technology companies do not roll out the same safeguards everywhere.

The same AI-generated advertisement might be blocked or labelled for a child in Europe and shown freely to a child in Kuching. That is not an accident of engineering. It is discrimination by design.

Need for caution

AI makes the old playbook more potent. Portions can be inflated, health claims invented, nutrition labels garbled.

Mascots can be created endlessly, each sliding past rules written for familiar, licensed characters. Ban one cartoon and a thousand are waiting behind it.

This matters because Malaysian children are not thriving. National health surveys show that close to three in 10 children aged five to 17 are overweight or obese, while stunting persists among the youngest.

To be fair, Malaysia has not been idle. The National Child Policy and Action Plan 2026-2030 puts digital safety at its centre, targeting grooming, cyberbullying, sexual exploitation and AI deepfakes.

But the digital threat it imagines is almost always a person: the predator, the bully, the abuser.

An algorithm that nudges a seven-year-old towards sugar every evening does not fit the picture, so it falls through the cracks.

A child’s health can be damaged by software as surely as by an individual, and the action plan ought to say so.

The Online Safety Act 2025 and existing rules on food health claims are a foundation, but neither was built for synthetic advertising.

Food-marketing rules need to be technology-neutral, and they should restrict techniques that appeal to children, not just named characters.

Platforms and advertisers should label AI-generated advertisements and switch on their strongest child-safety settings here, not only where regulators force their hand.

If a safeguard works in London, there is no technical reason it cannot work in Kuala Lumpur.

Let's come together

No single ministry can carry this. The government, alongside paediatricians, civil society and parents, must come together.

Malaysia should also push for an Asean standard. One country negotiating with a global platform has little leverage. Ten will have better chances.

Yes, children should learn that a beautiful picture is not always the truth but families cannot out-argue an algorithm alone.

Malaysian children are not second-class users of the internet, and it is time the companies profiting from them stopped treating them as if they were.

Writer is a lecturer affiliated with the Faculty of Business and Communication at Universiti Malaysia Perlis.

The views expressed here are those of the author/contributor and do not necessarily represent the views of Malaysiakini.