What this teenager wants you to know about the damaging effects of AI

CNN

29-01-2025

My friend Sammy turned around at his desk and grinned at me. He held out his phone and waved it in front of my face, the universal sign for 'Look at this!' It was a video of a capybara in a swimming pool, seemingly treading water like a human does, feet down and arms circling.
Last year, I would have believed the post I saw on that screen and accepted it as part of what I know to be true about the world: capybaras tread water like people.
Now, I wasn't sure.
Even though teens spend so much of our lives online, a new study by Common Sense Media found that teens between the ages of 13 and 18 increasingly do not trust the content they consume online.
With the emergence of generative AI — a type of artificial intelligence that produces content such as images, text and videos — it has become easy to quickly produce fake visual content. My friends and I have noticed these AI-generated images flooding social media platforms.
What's real and what's not?
Many teens struggle to figure out what online content is real and what is fake, according to the study. Some 46% said they either know that they have been misled by content or suspect that they have been, while 54% have seen visual content that 'was real, but misleading.'
Collected by Ipsos Public Affairs on behalf of Common Sense Media from March to May 2024, the survey data is from a nationally representative survey that includes responses from 1,045 American adults (age 18 or older) who are parents or guardians of one or more teens age 13 to 18, and responses from one of the teenage children in each of those parents' families.
Maybe the teens who reported that they haven't been misled are just absolute aces at spotting fake images, but I wonder if they just haven't realized that they have, in fact, been swindled by AI images.
'Many kids don't see flaws that they aren't trained to detect,' said Robbie Torney, Common Sense Media's senior director of AI programs, in an email. 'Knowing that the media we consume can be retouched, modified, or even fake, it's crucial that we all learn to think critically about information we encounter.'
Most teens have spotted deceptive content
I think that I'm generally pretty good at spotting AI images, but there was once a time when I believed that I couldn't be fooled. That was mostly in the early days of generative AI, when people in AI-generated photos typically had 15 fingers. But I no longer think I can't be fooled.
More than 70% of the surveyed teens who have experienced deceptive visual content reported it has changed how they view the accuracy of content. I'm part of that group now. After a few too many cool posts of alleged historical pictures that turned out to be fake, I generally distrust most pictures I see if they are not posted by someone I know. I examine the fingers of any human picture first, since I know AI can still get the number of human fingers wrong.
I have begun checking the comments of nearly every post I come across to see whether other users believe it. Doubt in the back of my mind now colors my response to everything I look at, and what I think.
Since third grade, I have been warned about checking sources and not believing everything you read. I have been generally distrustful of information from the internet for years, so AI has not suddenly damaged my trust in online information. What has changed is my trust in images. Seeing is not believing anymore.
I used to accept photos as fact. Before AI, there were programs people used to photoshop images, but those images are still real images of real people, just tweaked. Colors could be changed, and things could be cut out — but the original photo was an image of real life. As generative AI improves, however, completely fake content is getting harder for me to spot.
Many teens speak through images
One of the primary ways that my friends and I communicate is with images and videos. Memes are their own form of communication. Sharing an Instagram post about some weird scientific fact or interesting historical photo shows you know what a friend would find interesting. Sending someone funny video clips is a form of affection. It tells someone you remember them and lets you share an inside joke.
While I'm no longer a heavy social media user, I still probably see hundreds of images a day online. Other teens who spend more time on Instagram than I do or who use TikTok might see thousands. If those 'photos' are just images cobbled together by a machine, wholly detached from anything real, what can I believe? I already doubt everything I read online.
This mistrust is spreading into teens' offline lives. Studies show that teens already have low trust in institutions such as the government and news media. Doubt in the world around us is rampant among my friends. It's common for me to hear dismissive statements thrown around, even about the textbooks we read in class.
I've heard refrains like 'Eh, maybe they did that,' 'Might be true,' and 'I don't believe that' thrown out about anything and everything. My classmates and I often doubt the news, history and authority figures. If teens can no longer trust what they read, hear or see online, why should they trust anything?
In a generation in which mistrust could very well become the default approach to life, what can a person rely on? Or hope for? What's the point of caring about anything, if it could all be false? One of my teachers thinks that members of Generation Z (born between 1997 and 2012) are nihilists, whether they know it or not. My friends and I struggle with these questions, and I think my teacher may be right.
Generative AI needs a label
How does this get fixed? The Common Sense study found that 74% of teens agree that generative AI 'should have visible warnings that its outputs could be harmful, biased, or wrong.' And 73% of teens want content generated by AI 'to be labeled or watermarked' to show its source.
I believe this demand is because we're losing ground in what we can believe. I wish that I could trust what I see, and AI-generated content being required to be marked as fake is a reassuring idea.
'The growing mistrust in AI mirrors historical challenges with media literacy,' Torney said via email. 'Just as we learned to evaluate traditional media by asking 'who created this?' and why was it made?', we must now apply these same critical thinking skills to AI-generated content.'
It's a good solution, no doubt. But it still means that we cannot initially trust what we see. Instead, we must examine it, question it, and have faith in ourselves that we've come to the right conclusion.
People say that social media and increased time online is making us lonely because of the absence of physical connection. And maybe that's part of the deep loneliness, and the increasing anxiety and depression we know teenagers today report. But if this loss of trust continues, I think we'll become even more individualistic and form shallower connections with each other.
Mistrust should not be our default setting
If mistrust becomes the default approach to life, what would be the point of doing anything for other people that you don't already know? Will we be able to get to know each other deeply, or successfully communicate?
Can you really build a life when you don't know what is real and what is fake, when you can never trust what you see, what you learn, or how the world works?
Artificial intelligence is just that: artificial. When some of the main tools you use as a teenager to connect with real friends — memes and social media — are corrupted with artificiality, how do you make real connections? When the main way teenagers learn about the bigger world outside of their schools and towns is the internet, can we ever understand the world?