How can students evaluate AI content
The useful question is not “Was this made by AI?”
I have been thinking about how quickly “AI-generated” has become a kind of classroom shortcut. Students see a strange image, an unusually polished paragraph, or a video with a slightly wrong-looking face, and the first question is often whether artificial intelligence made it. That matters sometimes, but it is not the most useful place to begin.
The better question is: what evidence do we have for believing this, and what would change our minds?
That shift is at the centre of the Media Literacy Week 2026 materials from the National Association of Media Educators. The activities are designed for secondary students, but the underlying habits are useful far beyond one particular week in June. They can sit in Media Studies, English, Social Studies, or even a short form-time activity without needing a whole new unit.
AI is part of the problem, not the whole problem A generated image can be misleading, but so can a real photograph used without context. A genuine video can be clipped to remove the sentence that explains it. A human-written post can repeat an invented claim. A chart can contain accurate numbers while presenting them in a way that encourages the wrong conclusion.
If students are taught to detect only the visual signs of synthetic media, they may become overconfident. They might assume that a blurry image is fake, or that a professional-looking source must be trustworthy. Neither follows reliably.
I reckon the more durable skill is learning to separate the object from the claim. What exactly is this image, video, or paragraph asking me to believe? Who made it? Where did it first appear? What information is missing? Does another independent source support it? What might the creator gain from my attention, trust, anger, or money?
A small classroom routine One activity I would happily use is a slow comparison of two posts about the same supposed event. They do not need to involve a dramatic political story. A fake local weather warning, a manipulated celebrity quote, or a social media post about a school closure would be enough.
Students can mark each post in four colours:
- facts that can be checked - opinions or interpretations - missing context - details that create urgency or emotion
Then they can search for the earliest available version, check the date, look for an original image, and compare the wording with a reliable source. The point is not to reward the student who spots the fake fastest. It is to make the checking process visible and ordinary.
That last part matters. In real life, people often share something before they have time to investigate it properly. A classroom activity should not pretend that everyone can become a perfect verifier. It can, however, give students a pause button.
Creation changes the lesson The source materials also include the idea of responsible participation, which is important because students are not only audiences. They make videos, captions, memes, presentations, podcasts, and group chats. They choose what to crop, what to quote, and what to leave out.
A useful follow-up task would be to ask students to make a short misleading post using an entirely harmless subject, then annotate the techniques they used. Perhaps the school canteen menu, a fictional sports result, or an exaggerated review of a library book. The class can then revise each example so that the context is clearer and the claims are fairer.
That exercise usually makes media construction less mysterious. Misleading communication is not always a spectacular fabrication. Sometimes it is a true detail placed in the wrong frame.
A habit worth keeping after the themed week The best thing about the Media Literacy Week resources is that they do not treat media literacy as a one-off warning about scary technology. They connect access, analysis, evaluation, creation, and participation. Those are ordinary skills for living with media, whether the content was made by a person, a machine, or a mixture of both.
I am not sure students need another list of visual “tells” to memorise. They need practice asking careful questions when a post is designed to make them react immediately. That is slower, less glamorous, and probably much more useful.