z08: The Internet Is Becoming Impossible to Trust

 You see a video of a politician saying something shocking. You read a five star review for a product that looks perfect. You get a call from a family member, in a panic, begging for money. And all three could be fake. Not exaggerated. Not misleading. Fake, built by software in minutes, for almost nothing. The internet was supposed to be the greatest library humans ever made. Instead, it’s turning into a place where seeing is no longer believing. So how did we get here, and where does it end? Let’s follow the fakes.

Seeing Is No Longer Believing

In 2023, an image of the Pope in a huge white puffer jacket spread across social media. Millions of people believed it was real. It was made by a hobbyist with an AI image generator, and it was harmless fun. But that same year, a fake image of an explosion near the Pentagon went viral, and it briefly rattled the stock market before officials said it never happened.

Same technology, very different stakes. And the tools have only gotten better, cheaper, and easier since then. Today, anyone with a laptop and a few minutes can create a photo, a voice, or a video of events that never took place. No special skills, no studio, and no budget required.

The Scam Economy

The most immediate damage is financial. In 2024, an employee at the engineering firm Arup in Hong Kong joined a video call with what looked like the company’s chief financial officer and several colleagues. Every one of them was a deepfake. The employee transferred about twenty five million dollars before anyone realized.

Voice cloning makes it even more personal. Some tools need only a few seconds of someone’s voice to copy it convincingly. Scammers use that to call parents and grandparents, pretending to be a child in trouble. In early 2024, a fake robocall using an imitation of President Biden’s voice told New Hampshire voters to skip their primary. Regulators in the United States responded by ruling that AI generated voices in robocalls are illegal.

The Internet Full of Bots

Now zoom out. In 2024, a security company called Imperva reported that automated traffic made up about fifty one percent of all web traffic. For the first time in a decade, bots outnumbered humans online.

Then there’s what people call AI slop, mass produced low quality content, written and illustrated by machines, flooding search results and social feeds because it’s cheap to make and easy to monetize. Add fake accounts, fake followers, and fake reviews, and you get a strange new reality. Some of the crowd you see online isn’t a crowd at all.

Fake reviews became such a problem that in 2024 the Federal Trade Commission finalized a rule banning them, including reviews written by AI. Whether it can be enforced at scale is another question.

The Liar’s Dividend

Here’s the part that worries researchers most. Fakes are not the only danger. The existence of fakes is a danger too.

Legal scholars Danielle Citron and Robert Chesney named it the liar’s dividend. When everyone knows that anything can be faked, real evidence becomes easy to dismiss. A genuine recording of a politician, a leaked video of a company, a photo of a war crime, all of it can now be waved away with three words: “That’s AI generated.”

So trust doesn’t just fall for fake things. It falls for real ones too. And a society where nobody believes anything is just as easy to manipulate as one where everybody believes everything.

The Cost to Ordinary People

It’s easy to think this only affects politicians and big companies. It doesn’t. Fake investment ads use cloned faces of famous people to lure ordinary savers. Romance scammers use AI generated photos and chatbots to run relationships with dozens of victims at once. Students and workers are accused of cheating because unreliable detectors flagged their real work as machine written.

Every one of these cases chips away at something we rarely think about: the quiet, default trust that makes daily life work. We trust a stranger’s review, a news photo, a voice on the phone. When that default breaks, everything gets slower, more suspicious, and more expensive.

Even Search Is Getting Polluted

For years, a quick search was how we checked facts. Now results are crowded with pages written to rank, not to inform: product roundups nobody tested, health advice nobody reviewed, and answers copied from other machines. AI tools then learn from that same polluted web, so errors can be repeated, polished, and repeated again.

The old shortcut of trusting the first result no longer works. Finding reliable information now takes effort, and that effort is a hidden tax on everyone who just wants a straight answer.

Why the Platforms Struggle

You might ask why the big platforms don’t just fix this. Part of the answer is incentives. Algorithms are designed to keep you watching, and outrage keeps people watching. A shocking fake can travel around the world while a careful correction is still putting its shoes on.

The platforms are also changing how they police content. In 2025, Meta ended its third party fact checking program in the United States and moved toward a community notes system, where users add context to posts. Supporters say that’s more open and scalable. Critics worry it’s slower and easier to game. The honest answer is that it’s still being tested in real time, with real consequences.

Can Technology Fix It?

In fairness, people are building defenses. One approach is content credentials, an open standard from a group called C2PA, backed by companies like Adobe and Microsoft. It attaches a record to a photo or video showing where it came from and how it was edited. Another is invisible watermarking, like Google’s SynthID, which marks AI generated content.

But there are limits. Watermarks can be damaged or removed. Content credentials only help if cameras, apps, and platforms all support them, and many don’t yet. And AI detection tools are unreliable, sometimes flagging real images as fake and missing obvious fakes. Technology will help, but nobody should expect a magic button, and the fakers keep adapting.

The Other Side of the Story

It’s not all bad news. Human journalism, fact checking teams, and open communities like Wikipedia still do remarkable work. Many people are also getting more skeptical, which is healthy. Schools are teaching media literacy, and newsrooms are building verification desks. The goal isn’t to distrust everything. That’s the liar’s dividend again. The goal is to trust smarter.

What You Can Actually Do

You don’t need to quit the internet. But you can build a few habits.

First, pause before you share. Strong emotion is the fakers’ favorite tool. If something makes you furious or terrified in two seconds, slow down.

Second, check the source. Leave the page and search for who published it. Professional fact checkers call this lateral reading, and it works better than staring at the content itself.

Third, use reverse image search to see where a picture first appeared.

Fourth, protect your family. Pick a secret code word that only your relatives know, so if a panicked voice calls asking for money, you can verify it. And hang up and call the person back on their real number.

So, Can We Trust the Internet Again?

Maybe not in the old way, where seeing a video was proof. But trust can be rebuilt on better foundations: verified sources, clear labels, honest platforms, and people who think before they click.

The internet isn’t dying. It’s growing up, and growing up is uncomfortable. Every generation has had to learn to question what it was shown, and ours is simply next. Because in the end, the real question isn’t whether fakes will exist. They will. The question is whether we build a world where the truth is still easy to find.

If you want more stories about how technology is changing our lives, subscribe now and hit the notification bell. And tell me in the comments: what’s the most convincing fake you’ve ever seen online?

Thanks for watching!