I Had Social Media as a Kid. The Algorithms Are What Scare Me Now.

By Amber Ivey (The Human AI)

When I was in fifth grade in 1995, my “social media” consisted of AOL chat rooms where I hung out with friends after school. Through high school, I added BlackPlanet to the mix, and by the time I reached college, I eagerly joined MySpace and the college-only site then called TheFacebook. Teens have long found ways to carve out spaces to connect online, and that desire hasn’t changed.

The fundamental difference today is not that young people are connecting, but how they are being connected.

Back then, hopping online simply meant talking to people you actually knew in the real world. There was no endless feed, no infinite auto-play video loop, and no secret machine learning systems deciding what you should see. While pioneering psychological research in the late 1990s did begin tracking early forms of "Internet addiction" centered around AOL chat rooms, it was documented as a niche, isolated behavioral anomaly. The chronological layouts of the MySpace and early Facebook eras largely lacked the predatory architecture required to hijack a generation's biology. Users ultimately remained in control.

Today, that control has been stripped away. What used to be a casual pastime has been transformed into a carefully tuned dopamine machine.

Much of today's social media landscape is built around engagement-maximizing algorithms. AI has been woven into children's lives for years, quietly shaping the videos they watch, the content they encounter, and even the people they befriend. Unlike a TV episode, which has an ending, algorithmic feeds are designed to keep serving the next thing; modern social apps have no natural off-switch. Harvard Law expert Leah Plunkett perfectly captures this reality, comparing modern platforms to 24/7 entertainment venues where invisible robots repeatedly tap children on the shoulder and whisper that they should keep watching.

Medical research increasingly indicates that these precise algorithmic designs exploit developing brains. A 2025 literature review published in the journal Cureus notes that AI-driven feeds are systematically engineered to deepen the activation of the brain's reward centers.

When personalized content streams in, young brains respond by releasing dopamine, a neurological reaction mimicking the reward pathways associated with gambling or substance use.

This erosion of focus isn't anecdotal. A recent longitudinal study following more than 8,000 children found that higher social media use was linked to gradual increases in attention problems over time. Because notifications and endless scrolling encourage youth to switch their

attention quickly and often, they may gradually erode focus in a still-developing prefrontal cortex.

The mental health toll of this environment is equally stark. Meta-analyses reveal that for each additional hour teens spend on social media, the risk of depression jumps by about 13%. This is a design and systems question. In fact, youth themselves feel the manipulation: in a recent UK survey, 81% of older teens reported being acutely aware of recommendation algorithms in their apps. They know they are being managed by invisible forces.

Perhaps most chilling is how quickly these automated feeds can turn toxic when left to run unchecked. In a study by the Center for Countering Digital Hate, researchers set up test TikTok accounts for 13-year-olds. Within just 2.6 minutes, the system began recommending

suicide-related videos; within eight minutes, it was serving content promoting eating disorders. Crucially, accounts that signaled vulnerability through their usernames received 12 times more self-harm content than neutral accounts.

These are not abstract data points. They represent real-world tragedies, like that of 14-year-old Molly Russell, a UK teenager who died from an act of self-harm in 2017. A coroner later concluded that she died while suffering from depression and the negative effects of online content, and that the harmful material she encountered contributed to her death “in more than a minimal way.” In her final six months, Molly had reportedly viewed or engaged with roughly 2,100 posts related to suicide, self-harm, and depression. Children are being fed algorithmic poison on a loop, often outside parental sight or control.

It is vital to recognize that not all screen time is created equal, and our solutions must reflect that nuance. In the same study of 8,000 children, the hours kids spent watching traditional television or playing video games did not predict attention problems, only social media did. Television relies on fixed programming and commercial breaks, granting children a natural psychological exit ramp to turn the screen off and reclaim their agency. Algorithmic apps deliberately strip away that agency, operating on behavioral psychology principles shared by slot machines.

As policymakers increasingly debate sweeping smartphone and device bans or wholesale restrictions on social platforms, they must resist this reactionary overreach. Total device or app bans fail to respect modern youth culture, where social media connects kids with vital communities, identity support, and peer learning. Instead of treating the digital world as the ultimate enemy, lawmakers must focus on a far more surgical target: they must legally ban the deployment of algorithmic feeds and hyper-addictive design mechanics for minors.

Moving forward, our public policy and regulatory frameworks must focus on safety by design. The American Psychological Association explicitly recommends regulating features like endless scrolling and unrestricted feeds for minors. On the legislative side, we must demand frameworks like the "STAR" initiative, which urges lawmakers to force tech companies to build safety, transparency, accountability, and responsibility into their platforms. This means legally banning features like infinite auto-play for minors and forcing platforms to disclose exactly what content they are recommending to children.

We do not allow automakers to put cars on the road without seatbelts, airbags, and rigorous safety standards; we should not allow tech conglomerates to deploy cognitive traps for children without basic consumer protections.

Social media itself isn't a new phenomenon for kids; I am living proof that an entire generation grew up on it. What is entirely new is the predatory algorithmic attention economy built around it.

The fifth-grader I was got to log off and go do her homework. Today's kids don't get that off-switch, because somewhere along the way, their attention became too valuable to let them leave. The answer isn't to pull children out of the modern world. It's to stop letting that world be built to mine them. We can demand healthier design and hold these companies legally accountable. If we do, the next generation can keep connecting with their friends online, without being treated as lab rats for a profit-driven algorithm.

Amber Ivey is a data and AI governance practitioner focused on helping families and institutions navigate data and artificial intelligence responsibly.