It’s like a bad dream where the news you trust suddenly starts to feel…off. Imagine logging onto Facebook, scrolling through your feed, and seeing what looks like a New Zealand news page. It has the right logo, the familiar layout, and even uses stories you’ve seen on legitimate sites. But then you notice something unsettling in the images or videos accompanying the articles – something a little too perfect, a little too uncanny valley. This isn’t just a minor glitch; it’s a global problem, now hitting close to home in New Zealand, where a shadowy network of Facebook pages is churning out AI-generated content, twisting real news, and preying on our trust.
These aren’t just one-off incidents; investigations are finding a consistent, disturbing pattern. These fake news outlets lift actual articles from established, reputable sources like RNZ, the New Zealand Herald, and Stuff. They then slap on computer-generated images or short videos, often with slightly reworded text, and present it all as their own original reporting. It’s a clever, insidious tactic designed to trick unsuspecting readers. The examples unearthed in New Zealand are particularly jarring, but experts are quick to point out that this isn’t an isolated phenomenon. The same unsettling drama is playing out in news markets worldwide, often unnoticed by everyday people until the damage is already done, until misinformation has spread like wildfire.
The Australian Associated Press (AAP) did some digging and found one particularly egregious example: a page called “NZ News Hub.” This page was a prolific republisher, taking stories straight from legitimate New Zealand news sources. But here’s the sinister twist: they didn’t just copy the text. They overlaid it with AI-produced images and short videos, making it all look like original content. The page’s “About” section would claim to offer “latest New Zealand news, breaking stories, politics, business, sport, and community updates.” And it worked. Thousands of people followed it, and it consistently generated engagement – likes, comments, shares – despite creating absolutely no real journalism itself. It was a phantom news organization, thriving on the back of others’ legitimate work.
Sometimes, the cruelty of this practice becomes shockingly apparent. Consider the tragic Mount Maunganui landslide, which claimed six lives. In a horrific display of disrespect, a still photograph of a 15-year-old victim, Sharon Maccanico, provided by the police, was animated to make it appear as if she were dancing. RNZ, the actual news outlet that would have covered such an event, confirmed unequivocally that no such video was ever recorded by their crews. Further fact checks by the AAP and others revealed a chilling truth: many of the images linked to the disaster were riddled with errors. They showed geographical inconsistencies, depicted implausible details, or even carried digital watermarks that most users wouldn’t recognize, all clearly indicating they were generated by AI, not captured at the scene. It’s hard to imagine a more callous use of technology than to exploit human tragedy for clicks.
So, why are these digital puppeteers doing this? Experts like Andrew Lensen, a senior lecturer in AI at Victoria University of Wellington, explain the motivation is brutally straightforward: “These pages want to get as much engagement (reactions, comments, shares) as possible, in order to build their following/exposure and potential ad revenue.” It’s all about the money and the reach, the digital equivalent of a con artist selling snake oil to a desperate crowd. The explosion of easy-to-use generative AI tools has lowered the barrier to entry, making it simple for almost anyone to create what looks like a professional news operation. Lensen even points out that some synthetic images carry subtle watermarks, like Google’s SynthID, which most users wouldn’t even recognize as a sign of AI generation. This ease of creation, combined with the lure of engagement, fuels this unsettling phenomenon.
This isn’t just one or two rogue pages; it’s a systemic problem. Other news organizations have also documented this shadowy behavior. A 1News analysis, for instance, identified at least ten separate Facebook pages that were all repurposing local reporting, running it through generative AI systems, and then publishing it with fabricated visuals. One review found that a single page managed to post over 200 items in a mere month. Beyond this, separate AAP fact checks have uncovered repeated instances where supposed footage of politicians, police responses, or even grieving families was completely fabricated or manipulated. And it’s not confined to Facebook either. Fact-checking organizations report that these deceptive images and clips are popping up on TikTok, Instagram, and X within minutes of major breaking events, spreading misinformation rapidly across various platforms. The true puppeteers behind these operations are often elusive; transparency data for Facebook pages shows that many of these accounts are administered from overseas, with operators in places like Vietnam and Malaysia. This geographic distance complicates any attempts to understand their true intentions or hold them accountable. Even when platforms manage to shut down these fake pages, moderators report that near-identical clones often reappear almost instantly, like hydra heads sprouting anew.
The legal landscape offers little comfort against this digital onslaught. New Zealand’s Classification Office states that the law treats AI-generated material no differently from any other content under the Films, Videos, and Publications Classification Act 1993. What matters for legal purposes is what is depicted, not how it was created. This means that if an AI-generated image shows harm, it can be illegal, but the act of AI generation itself isn’t necessarily against the law. This legal loophole leaves a lot of room for exploitation. Even civil defense agencies and community groups are now issuing public warnings about synthetic posts during emergencies, recognizing the very real-world harm that misinformation can cause, especially when lives are on the line.
Mainstream news outlets are understandably treading carefully in response. RNZ, for example, has published its AI principles, stating that it will generally not knowingly disseminate content created by generative systems. This cautious approach reflects a growing awareness of the dangers. Some industry observers believe that this crisis of misinformation might actually lend renewed authority to trusted media, as people seek out verified information from reliable sources. However, others warn that if legacy organizations themselves start relying too heavily on AI for content creation, they risk further blurring the already murky lines between verifiable fact and dangerous fabrication. The challenge for news organizations is immense: how to leverage powerful new technologies while upholding the fundamental principles of truth and accuracy in a world increasingly flooded with digital illusions.