Older Internet Users Increasingly Struggle to Spot AI-Generated Images

A picture of Jesus made of seashells, a dog made of noodles, or the pope made of potatoes might sound like a joke from an internet archive, but it's the daily reality on millions of Facebook accounts. A new study finds that the older an internet user is, the harder it becomes to tell such images apart from real photos, and the creators of this content are deliberately targeting older generations.

What the study found

The authors of "Human Factors in Detecting AI-Generated Portraits: Age, Sex, Device, and Confidence" tested 1,664 people aged 20 to 69, asking them to identify which of the portraits shown were real and which were generated by artificial intelligence. Average accuracy came to 85.2 percent, but the score consistently worsened with age. In other words, the older the test participant, the greater the chance they mistook a fabricated image for a real photo. This isn't just an academic concern, the results line up with what social media analysts have been observing.

Seashell Jesus and other slop hits

The nickname "seashell Jesus" refers to a specific type of image that regularly floods Polish and international social feeds: religious figures or animals assembled from unusual materials such as seashells, sand, or vegetables. Alongside seashell Jesus, the same category includes sand Jesus, a noodle dog, and a potato pope.

An analysis published in the HKS Misinformation Review covered 125 such Facebook pages and found that the average one has around 147,000 followers. One image, showing a fancifully decorated kitchen, racked up 40 million views and nearly 2 million reactions, comments, and shares combined. The scale of those numbers shows that slop is no longer just an internet curiosity, it's become a real channel for reaching millions of people.

Why it works on older generations

Oskar Barcz, an IT specialist and vice president of the Generacja Innowacja Foundation, tells android.com.pl that the problem doesn't lie solely in older internet users' lack of technical skills. In his view, the key issue is that detection tools simply can't keep pace with how fast image generators are improving.

There's no single effective tool for spotting AI or photo manipulation, because the technologies used to create them change week to week - Oskar Barcz, vice president of the Generacja Innowacja Foundation

Barcz points out that knowing how to spot specific visual artifacts isn't enough on its own, since image generators improve at a pace that's hard to catch up with through one-off training. What matters more is building general vigilance toward content landing in an inbox or on a Facebook feed. At the same time, the expert warns against swinging to the other extreme, total distrust of everything seen online.

That kind of attitude leads to fear, isolation, and distrust even of genuine information. What matters is something different: teaching people that content online, including a video recording or the voice of a loved one, is no longer proof in itself these days - Oskar Barcz, vice president of the Generacja Innowacja Foundation

What the AI Act changes

Since August 2, 2026, the EU's AI Act has imposed new transparency obligations on creators and platforms regarding AI-generated content. Barcz cautions, however, that the new law doesn't solve the problem on its own, since enforcing labeling requirements across millions of accounts producing slop remains a daunting task for regulators. In practice, this means the responsibility for recognizing fake content still largely falls on users themselves, especially those who, as the study shows, handle it the worst.

The law increases accountability for platforms and tool developers, but it doesn't relieve us of the need for caution - Oskar Barcz, vice president of the Generacja Innowacja Foundation

Where to find help

Barcz notes that more and more local institutions are recognizing the scale of the problem and organizing free digital literacy classes aimed specifically at seniors. He points to senior clubs, community centers, libraries, and universities of the third age as places where people can learn basic online safety habits. For Polish families, the practical takeaway is simple: it's worth talking with older relatives about what they see on social media, before one of these images gets used as a pretext for stealing data or money.

Senior clubs, community centers, libraries, universities of the third age, and community organizations are increasingly organizing free meetings with experts - Oskar Barcz, vice president of the Generacja Innowacja Foundation

As the authors of the HKS Misinformation Review analysis note, slop pages build up their reach precisely so they can later steer followers toward ad-filled sites or exploit the trust they've built for scams and data theft.