How fake news pages work, and why spotting one isn’t the end of it

Could you tell a real news page from a fake one built to farm your clicks? During the Thailand-Cambodia conflict, at least 50 pages profited from the outrage they stirred – here’s how they work, and how to spot them.

By Ken G Weber and Nicha Wachpanich

Published 6 July 2026
5 min read

The audio sounded like the Thai Prime Minister. After closing the border crossing between Thailand and Cambodia last year, the short TikTok clip suggested Anutin Charnvirakul planned on reopening it in June.

But it was not the Prime Minister talking. The audio was fabricated by AI – verified as fake by fact-checkers from Thai PBS and by Anutin himself.

As of today, Thailand has not reopened the border with Cambodia.

The AI-generated clip spread from platform to platform, and with each repost it looked more like news. Similar misinformation proliferated last year during the armed border clashes that killed dozens and displaced hundreds of thousands, a confrontation that quickly spilled into cyberspace.

Last year’s clashes have cooled, but the machinery built to exploit them has not been dismantled. When Thais headed to the polls earlier this year, the same infrastructure was waiting. A HaRDstories investigation found at least 50 pages exposing millions of Thai netizens to false evacuation orders, distorted death tolls and other misinformation. Operated by admins in Cambodia, the pages point to information operations (IOs) that experts say can be hired out, functioning as PR companies for anyone with the necessary funds and malicious intent.

“These are commercial accounts. Their goal is to make money,” said Jerry Yu, a senior analyst at Doublethink Lab, a Taiwan-based research organisation that monitors information operations. “They don’t choose buyers. When conflict happens, it’s the time for them to make money.”

When fact-checkers are overwhelmed by information dumps, verifying and detecting the latest falsified content takes time. So how do we to distinguish reliable news sites from imitations?

How fake pages operate

At first glance, “Daily News 24” – a page now removed from Facebook – looked innocuous: weather forecasts, car crashes and Buddhist teachings. But dig a little deeper and the page became a portal to gambling sites.

Thais might find the name familiar: it echoes the country’s second-largest newspaper, Daily News. These fake sites often mimic an established outlet, borrowing its name and inflating follower numbers to manufacture legitimacy, according to DISARM, an open-source disinformation framework.

The pieces work together. Everyday posts make the page look like an ordinary news source. Nationalist and sensational content, increasingly generated by AI tools, works as the hook that builds an audience. Gambling links and scams are how that audience is cashed out. The innocuous material builds trust, engagement drives reach, and monetisation follows – which is why, during the border conflict, war content proved so valuable.

A recent investigation by 101.World found AI-generated content received high engagement rates.

In one such post, a photorealistic Thai soldier stands on a battlefield and asks viewers for “just one share”, not money, not belief, just the share. Highly nationalist content collects views and likes during the conflict – and, with them, the ad revenue that keeps the pages running.

Fake news pages run from Cambodia target Thailand. But who’s really behind them?

by Nicha Wachpanich

•

How fake news is turning Thais against Myanmar migrants

by HaRDstories

“In a combat situation like this, people may not care so much whether something is true or false,” said Senator Tewarit Maneechai, a former journalist, during a recent discussion on disinformation campaigns in Bangkok. He described how such content feeds on nationalist sentiment, turning mockery of the enemy into entertainment.

Who is behind them?

HaRDstories traced at least 50 pages back to Cambodia through Meta’s Page Transparency feature, but the trail stops there: the tool reveals a page’s location, not its operator. Yu, of Doublethink Lab, calls this a “butterfly attack,” after a butterfly that mimics another species’ flight pattern. The actors, he said, may be impersonating a Cambodian operation to hide their real origin.

“As some content attacks political parties, the assumption could be that the threat actor is actually a competing party. They want to pretend it’s an attack from Cambodia,” said Yu.

Yu’s Doublethink Lab published a report on an information operation targeting Taiwanese and Japanese internet users. Tracing a large, interconnected network of content run by a Chinese content farm, the team concluded that the Chinese government was behind it.

During Taiwan’s 2024 election, the lab found, fake Taiwanese accounts operated from Cambodia sought to sway voters by using AI-generated news anchors and coordinated attacks.

Another clue to the administrators’ identity may be the pages’ location: Cambodia. With scam operations across the country running at industrial scale, it is not a stretch to see disinformation as part of the same ecosystem.

“Most cases documented so far show the fine line between scams and disinformation,” said Benedikt Hofmann, deputy regional representative of the UN Office on Drugs and Crime (UNODC). “The same technology common in scams can easily be directed towards targeted disinformation campaigns as another income source for criminals behind these centres.”

A Meta threat report from 2025 points to posts on the platform that are linked to scamming operations, posting content impersonating government pages including those from Thailand. Meta has removed over 6,400 accounts on Facebook relating to scam networks, according to the report.

What platforms are doing to combat this

The 50 pages investigated by HaRDstories fall into distinct networks that in some cases produce the same exact content posted across different pages. This violates Meta’s Community Guidelines on coordinated inauthentic behaviour, where multiple accounts operated by an individual or group deceive and expose users on the platform to falsified content.

HaRDstories spoke to a Meta representative about the company’s efforts. Meta’s recent threat report acknowledged the networks were active during the conflict, but the company would not say who was behind them, or whether it is still monitoring the specific pages HaRDstories asked about.

After HaRDstories alerted Meta, the pages under investigation were removed from the platform.

A Reuters special report last year found that Meta had projected about 10 percent of its 2024 revenue – roughly 16 billion – would come from ads for scams and banned goods.

Why spotting fakes isn’t enough

A recurring tactic seen in fake news sites is the usage of AI-generated content. Awkward language and hyper-realistic images are obvious clues and AI detection software can help flag inauthentic content posing as news.

Users can also check a page’s Page Transparency and ad details to see who manages it and from what country. The ad section can likewise reveal what the page is really promoting.

But spotting the fakery is not the same as being immune to it. In an investigation into AI-generated influence networks, 101.World spoke to Kulthida Samaputthi, an editor at the fact-checking organisation Cofact, who cautioned against assuming that an obvious fake is a harmless one.

Content that one person dismisses as AI can still persuade someone else, she told 101.World. Audiences are not uniform and the danger builds through repetition. “Disinformation works by getting us to see it over and over, from many places,” she said. “Even if we recognise at first that something is AI, over time, or as the context shifts, memory can blur until we take it for real.”

Nattakorn Plodde, digital verification editor at AFP for six years, views fake news as an ‘infodemic’ as misinformation spreads rapidly during times of crisis.

“If we cannot manage falsehoods, then ten years from now people will see the same event through two different lenses,” he said. “What will happen to society then?”

This explainer was supported by Internews

Ken G Weber is a Thai-Austrian journalist who reports on overlooked issues in Thailand. His writings are inspired by local voices seeking to defend their rights and environment.

Nicha Wachpanich is a Bangkok-based journalist covering stories from a rights-based and human-interest perspective. She previously worked with a local environmental news agency under Thai Society of Environmental Journalists.