Deepfakes PMX !?!
Have you ever come across those too-good-to-be-true videos? You know, the ones that scream "get rich quick," promising fast money with minimal effort? I’m just mashing words here, but you get the point. Back in the day, these scams came in the form of shady emails or spammy text messages. But now, with AI crawling into every corner of our digital lives, these scammers have evolved.
They’re using AI to slide into your WhatsApp chats or flood your TikTok feed with a strangely familiar voice, a well-known face, speaking fluent Bahasa, encouraging you to invest. “Sedagha-sedaghi, kita labur hari ini, esokkk kitaa menang !!.” Even I your 24/7 IT guy glued to the internet struggle to spot the difference sometimes. Now imagine the Gen X crowd, boomers, or even millennials who aren’t as tech-savvy.
PMX Telling Us to Invest More or Is He...?
In case you’ve been living under a rock (no shame, hermit crabs are cool too), Malaysia is currently under digital attack. Scammers, cybercriminals, and even politically motivated actors are using AI to mimic the voices and faces of Malaysia’s most prominent figures. We're talking about our very own beloved PMX. But it doesn’t stop there. Influencers like Khairul Aming and even the royal family have been deepfaked to promote false narratives, scam schemes, and stir public unrest.
Don't trust me ? no worries,
According to MCMC via FMT reporter:
Statistic On Trend Cases of The Usage of AI
That’s just for AI-manipulated content. If we’re talking fraud overall, check this stat from Bernama and Marketing-Interactive:
Statistic On Trend Cases of Fraudulent using AI
The fraudulent online content involved the use of artificial intelligence (AI) such as deepfakes, impersonation of notable figures, videos, graphics and text.
Examples? Oh, there are plenty:
Those are some serious number that are surging up, and are still on the rise. Some of the example of these AI manipulative content:
- A deepfake video of PMX endorsing a fake investment scheme.
PMX is Offering Us Easy Money or Is He ...?
- An AI-generated voice of PMX and Elon Musk promoting cryptocurrency fraud.
AI Anwar Talking to Fake Elon Musk
- A viral AI-generated image claiming illegal mining in a Raub durian farm, which amassed over 100,000 views before being debunked.
AI Image of The So Called Raub Gold Mine
Now you’re probably wondering: Is this even illegal under speech law?
I mean, who are you gonna blame? The robot? The software? The machinery?
Valid question. Tracking down the culprit feels like chasing Leonardo DiCaprio in Catch Me If You Can. But legally? These acts are just as illegal as if you yourself had made the fake account to impersonate those high ranked profile characters.
Because under the Communications and Multimedia Act 1998 (CMA):
- Section 233(1)(a) criminalizes the distribution of false, indecent, or menacing content online intended to harass or deceive. which is literally deepfaking someone for your needs.
Still not enough? How about the Penal Code:
- Section 416 – Cheating by personation, which can now include AI impersonations.
- Section 499 – Criminal defamation. Use AI to damage someone’s reputation? Congrats, you're legally toast.
Okay, let’s say those laws aren’t enough if you’re caught doing this, you could also be charged under federal law, specifically the Sedition Act 1948:
- Section 3(1) defines seditious acts as anything that promotes contempt or disaffection against the government or monarchy. So, if your AI makes the YDPA say something false, that’s a full-on criminal offense.
Look these laws may be decades old, but don't underestimate the consequences of going against them.
But what about freedom of speech whaaaaaa?
Yes, I can already hear the SJWs smashing their keyboards arguing that this kind of control pulls us backwards as a society! Or that not using AI is like cutting off our own legs because AI is the future, so why not explore it now? Or the classic cry: “What about freedom of speech?”
Look, this isn’t about banning AI or shutting down citizen voices. It’s about understanding that actions do have consequences. Yes, Malaysia guarantees freedom of speech under Article 10 of the Federal Constitution, but it is not absolute. This is qualified freedom meaning it’s subject to restrictions in the interest of national security, public order, or morality.
Creating AI deepfakes of national leaders to scam or troll? That’s not protected speech. That’s deception, and it comes with serious consequences.
The Law Can’t Keep Up
AI tools like ChatGPT, Gemini, ElevenLabs, DeepFaceLab, and OpenAI’s Sora are evolving rapidly. By the time our lawmakers attempt to combat these acts, a dozen new ways to fake reality have already emerged. In Malaysia especially, we lack an AI-specific legal framework. Don’t believe me? No problem trust the word of a law firm established in 2010 with just ten lawyers, which has since grown to become one of the largest corporate law firms in Malaysia, with over 90 lawyers based in Kuala Lumpur.
Rahmat & Lim states, “There is presently no specific law regulating the use of AI in Malaysia.”
They also touch upon a guideline outlining seven key principles referred to as the "AI Principles" which will form the basis for AI development.
National Guidelines On Artificial Intelligence Governance and Ethics: Key Principles
My message to the government current or future:
Take this matter seriously. This is the same mistake made by students in the early days of social media they said whatever they felt like, even if it was slanderous, harsh, or intended to ruin someone’s reputation. And in the end, many faced the consequences they deserved.
We need to learn from past mistakes and look ahead. From where I stand, AI is the future. Our nation’s lawmakers need to stay in touch with this issue and understand the urgent need for AI-specific legislation laws that clearly define and outlaw malicious deepfakes.
Furthermore, platforms must be held accountable. Tech companies need to detect and remove deepfakes more efficiently. And finally, we, the people, must do our part by raising public awareness especially among older users so they understand that not every familiar voice or face online is real, it all goes back to the all saying "don't trust everything you see on the internet".
I'll end you with this video, explaining some of the tactics used in this type of AI attack:
If it looks like PMX and sounds like PMX, but he’s offering you a 300% return on bitcoin or NFT, it's probably not him.
Source
- https://www.freemalaysiatoday.com/category/nation/2024/12/04/mcmc-records-huge-increase-in-explicit-ai-generated-content
- https://www.marketing-interactive.com/fraudulent-content-removed-social-media-2024
- https://www.bernama.com/en/news.php?id=2379329
- https://www.rahmatlim.com/perspectives/articles/29210/mykh-national-guidelines-on-artificial-intelligence-governance-and-ethics-key-principles#:~:text=There%20is%20presently%20no%20specific,use%20of%20AI%20in%20Malaysia.
- https://www.suarakeadilan.my/post/lombong-emas-di-balik-kebun-durian-di-raub-palsu-janaan-ai
- https://x.com/biarbetul_rasmi/status/1824055127428530469
- https://x.com/biarbetul_rasmi/status/1874076771928465895
- https://x.com/biarbetul_rasmi/status/1799051698176196918
Written by Limau Ais (Amal Akmal)
No comments:
Post a Comment