Contradictheory: Is the Internet dead?

Bot-created content made up 51% of all Internet traffic in 2025, meaning that humans are now the minority species online, says the columnist. — 123rf

“On the Internet, nobody knows you’re a dog” is something I like to say from time to time. Of course, these days I’m more likely to be ranting at some youngster who was born after this saying was coined, and I have to explain to them what exactly it means: On the Internet, you can say you’re anyone and claim anything, because nobody knows who you are. 

Actually, these days, I should say, “On the Internet, nobody knows you’re a supercomputer powered by thousands of GPUs”. Less catchy, but more accurate.

Even before the advent of publicly accessible artificial intelligence (AI), it was posited that more than half the content on the Internet was not written by humans. Indeed, according to cybersecurity firm Imperva, bot-created content made up 51% of all Internet traffic in 2025, meaning that humans are now the minority species online.

It should not be too surprising. Most of this bot traffic is a result of automated tasks shuffling back and forth across the Internet, as search engines, smart devices, and automated scripts do what they do to keep the Internet running.

Yet this is also the basis of something called the “dead Internet theory” that claims most of what we see day to day on the Internet is generated by bots and algorithms. This includes social media posts, political blogs, and even the passionate comments arguing below them.

The most extreme version of this is that it’s a conspiracy by governments and rich private corporations who want to manipulate the masses to do their bidding.

On the one hand, it shouldn’t be surprising that we as humans are repurposing and resharing existing human knowledge. It’s what we do almost every time we talk to other people. Probably 80% of what you’re reading in this column is extracted from articles and studies written by other people.

Does it matter if it’s a human being or a computer generating this?

As usual, it depends on the context. Are you trying to sell me something? Or persuade me not to vaccinate my kids? It matters very much.

A 2024 article in the British Medical Journal (BMJ) pointed out that misinformation about vaccines spread so quickly online that it outpaced public health countermeasures. The authors argued that social media platforms needed to take greater responsibility for the accuracy of what appeared on their feeds.

In Malaysia, it’s the government that’s been trying to play this role of responsible curator. And truth be told, we’ve been enthusiastic about this. In 2018, the government under then prime minister Datuk Seri Najib Razak introduced the Anti-Fake News Act after claiming that the Opposition was using fake news to win over voters, including on social media sites.

It was repealed the following year under a new government, but then in late 2024, Communications Minister Fahmi Fadzil touted the need to license social media platforms to address criminal activities, especially those targeting children.

I hope you recognise the blatant irony that the push to decide what you can and can’t see on the Internet might be, in fact, determined by governments and rich private corporations.

The thing is, I do think that the Internet is full of scams, fraud, and hate speech, and governments should regulate content that may be harmful to society.

Bank Negara Malaysia’s Financial Consumer Alert List is a helpful example of how to handle it: identify unlicensed investment schemes, make the information public, and let users decide for themselves. The National Scam Resource Centre completes a feedback loop with the public by offering a hotline for victims. These are practical, targeted solutions. What they are not are blanket bans.

Because efforts to just shut off parts of the Internet may backfire. The BMJ article noted that suppressing misinformation often pushes it underground, where it festers unchallenged in closed groups and private channels. It’s hard to chop down the weeds when you can’t see where they are.

That’s also why it’s not a good idea to ban under-16s from using social media, as floated by the government earlier this year. Nothing makes bad ideas more attractive than a government warning label.

Ultimately, it feels like any solution to regulating the Internet will be a complex one that sits somewhere between chaos and control. While there is an overwhelming volume of misinformation that can now be quickly generated by AI, the solution cannot be governments and corporations positioning themselves as arbiters of truth, using archaic tools that may be blunt, inconsistent or, perhaps worst of all, politically convenient.

You need to address wilful abuses and untruths, but we also need to maintain a freedom to express all sorts of ideas. Accountability should not mean invasive surveillance, regulation should not mean arbitrary censorship. And “content moderation” should not quietly become “truth management”.

Once upon a time, the Internet was a place where anyone, even a dog, could say what they thought and it was heard around the world. That asymmetry gave the web its power to let the small and the obscure be seen and heard.

Now the powers that be are trying to sanitise what they see as the world wild web. They would prefer no more fake news, or offensive posts that disinform. Instead, they would be happy with AI-generated politeness, approved by the relevant ministries, targeted at the right electoral demographics.

When that day comes, I hope there’s still at least one hound left typing furiously on a keyboard, to remind everyone that once upon a time, the Internet was alive.

In his fortnightly column Contradictheory, mathematician-turned-scriptwriter Dzof Azmi explores the theory that logic is the antithesis of emotion but people need both to make sense of life’s vagaries and contradictions. Write to Dzof at [e-mail]. The views expressed here are entirely the writer's own.