A country can survive disagreement. What it struggles to survive is manufactured hatred dressed up as public opinion.
When Malaysia’s Deputy National Unity Minister recently claimed that 70% of online comments surrounding the “kuil haram” controversy were AI-generated, it triggered both concern and confusion.
Concern, because if true, it suggests technology is now being used to inflame one of Malaysia’s most sensitive fault lines: race and religion.
Confusion, because it risks making the machine the villain while the human hand behind the keyboard disappears.
That would be a dangerous mistake.
Because the greater threat is not artificial intelligence.
It is weaponised human intelligence.
Because even if the comments were generated by artificial intelligence, AI did not choose the target. AI did not decide which temple issue to inflame. AI did not wake up angry one morning and decide to attack a community.
People did.
And that distinction matters.
The Global Playbook Arrives Locally
Around the world, troll farms and bot networks have already shown how digital manipulation works. From election interference campaigns in the United States and Europe, to coordinated disinformation networks in the Philippines, India, and elsewhere, operators have used fake accounts, emotional messaging, and algorithmic amplification to distort public perception. Malaysia should not assume it is immune.
What begins overseas as a tactic often arrives locally as a template.
The New Face of an Old Game
Malaysia is no stranger to the politics of 3R Race, Religion, and Royalty.
For decades, these themes have been used to mobilise fear, distract from governance failures, and fracture public unity. What changes over time is not the method of manipulation, but the tools available.
Yesterday, it was pamphlets, whispers, chain messages, and anonymous blogs.
Today, it may be fake profiles, coordinated comment farms, edited videos, bot networks, and AI-generated text.
The technology is new.
The motive is old.
AI Does Not Hate on Its Own
There is an important reality many Malaysians should understand.
Mainstream AI systems usually include safety filters that reject direct hate speech, racial slurs, and content designed to insult communities. Ask a public AI tool to rewrite a racist insult and it will often refuse.
For example, if someone asks an AI to rewrite a statement insulting Malay based on skin colour or hygiene stereotypes, most responsible systems will reject the request and redirect the user toward respectful language.
That means when we see highly targeted hateful comments online, we should ask a smarter question:
Was this really spontaneous AI behaviour?
Or was it human intent using technology as camouflage?
Because machines can generate text.
But humans generate agendas.
Coordinated Inauthentic Behaviour
The more accurate danger may not be “AI comments” in the abstract.
It may be coordinated inauthentic behaviour.
This happens when operators create multiple fake accounts, flood comment sections, repeat emotionally charged narratives, and manufacture the illusion that hatred is widespread and organic.
One person can run ten accounts.
One team can run hundreds.
One script can produce endless variations.
Suddenly, an ordinary citizen scrolling through Facebook or Instagram sees what appears to be mass outrage.
And perception becomes reality.
That is how digital tension spreads.
Not because everyone is angry.
But because someone wants it to look that way.
Why the “Kuil Haram” Issue Is So Easily Exploited
Few phrases are more combustible in Malaysia than those involving houses of worship, land disputes, ethnicity, or historical grievance.
A temple controversy is not just a local planning issue once it enters social media.
It becomes symbolic.
One side sees legality.
Another sees dignity.
Another sees selective enforcement.
Another sees political provocation.
That complexity creates the perfect environment for manipulators.
All they need to do is push the most inflammatory interpretation.
And the algorithm does the rest.
Human Direction Still Sits at the Centre
Even the most advanced tool still needs instruction.
Someone chooses the temple to target.
Someone decides which phrase to weaponise.
Someone writes prompts designed to bypass safeguards.
Someone selects the fake profile photo.
Someone schedules the posting time.
Someone benefits from the outrage.
That is why blaming “AI” alone can become convenient.
It shifts attention away from the organisers.
The keyboard may be automated.
But the intent is manual.
Do Not Let Technology Become a Shield
There is another danger in overusing the AI explanation.
If every toxic campaign is blamed on bots or machines, real provocateurs gain cover.
They become invisible behind jargon.
Instead of asking who funded it, who coordinated it, who owned the accounts, and who gained politically, we end up debating software.
That serves the wrong people.
Malaysia does not need only digital literacy.
It needs accountability literacy.
Enforcement Must Catch Up
If authorities genuinely believe fake networks are exploiting 3R issues, then action cannot stop at press statements.
There must be faster investigation into:
- Fake Facebook and Instagram accounts
- Repeat coordinated posting patterns
- Paid influence networks
- Cross-platform harassment campaigns
- Identity misuse and impersonation
- Deliberate incitement of racial hatred
The rakyat do not need another warning headline.
They need visible consequences.
Malaysia already shows urgency in other digital harms. Regulators have highlighted thousands of takedown requests involving scams and fraudulent content, often with rapid coordination. That raises a fair public question: if scam posts can be pursued at scale, why do racial insults, community hatred, and coordinated offensive pages often appear slower to address?
MCMC often acts as an intermediary, flagging content and requesting removals from platforms such as Meta and TikTok, which then make final decisions under their own terms of service. That makes platform cooperation essential but it also means enforcement pressure must be equally serious when national harmony is at stake.
Existing communications laws and provisions against incitement should be applied carefully, fairly, and without selective hesitation.
Because impunity is the best software for future abuse.
The Public Also Has a Role
Ordinary Malaysians must also become harder to manipulate.
Before sharing outrage, ask:
- Is this a real person or a suspicious account?
- Is the language designed to provoke emotion instantly?
- Is this issue being selectively framed?
- Why is this appearing now?
- Who benefits if communities start fighting?
That pause matters.
Many digital fires survive because citizens provide the oxygen.
Malaysia’s Real Test
Malaysia has always managed diversity imperfectly, but often peacefully.
That balance is precious.
It should not be surrendered to anonymous avatars, cynical operators, or imported culture-war tactics.
The real challenge is not whether AI can write hateful comments.
It is whether Malaysians will allow themselves to be steered by them.
Closing Reflection
The recent claim about AI-generated comments may be partly true, partly exaggerated, or somewhere in between.
But one truth is beyond dispute:
No algorithm filed the grievance.
No software felt resentment.
No chatbot planned the division.
People did.
So by all means regulate technology.
But do not forget to pursue the humans hiding behind it.
Because if Malaysia mistakes the tool for the culprit, the real architects of division will keep working quietly in the dark.
And the next keyboard riot is only one trending topic away.
If scams threaten wallets and hate threatens the nation, why is only one treated urgently?
Annan Vaithegi, on truth and manipulation.
Annan Vaithegi (annanvaithegi@icloud.com) is a content creator under the Newswav Creator programme, where you get to express yourself, be a citizen journalist, and at the same time monetize your content & reach millions of users on Newswav. Log in to creator.newswav.com and become a Newswav Creator now!
The User Content (as defined on Newswav Terms of Use) above including the views expressed and media (pictures, videos, citations etc) were submitted & posted by the author. Newswav is solely an aggregation platform that hosts the User Content. If you have any questions about the content, copyright or other issues of the work, please contact creator@newswav.com.