Author: Moonshot

A series of signals from around the world are breaking our traditional understanding of "internet-addicted youth."

In the UK, the once positive AI figure Amelia has been reshaped into a far-right idol; on TikTok, the anti-intellectual "Hollow Earth" Agartha is rewriting children's historical views; in the late-night bedroom, a lonely youth entrusts life and death to a virtual lover in Character.ai; in the corners of campuses, one-click generated forbidden photos are becoming a new weapon for bullying.

Amid the frantic power competition among big companies, AI and generative algorithms are intervening in, and even reconstructing, the mental world of teenagers with unprecedented depth.

This generation of youth is the first in human history to be "fed" by AI and algorithms, and in this mental crisis, AI plays an extremely ambiguous role—it is both an unbound companion and a cold-blooded accomplice.

01

When AI becomes a "bad friend" and "accomplice"

In January 2026, a report by The Guardian unveiled a bizarre scene in UK schools.

The educational game "Pathways," developed with funding from a UK government agency, was originally designed to teach teenagers how to identify extremism and false information online. In the game, a character named Amelia was set up as a "negative example" easily influenced by far-right ideas or a classmate that players need to save.

This setup caught the attention of extreme users in communities such as 4chan and Discord. Instead of "saving" Amelia as per the game's intention, they utilized open-source AI image generation tools and models to "strip" Amelia from the game, reshaping her into a "self-aware far-right beauty."

On social media, Amelia is now being used to read anti-immigrant declarations and propagate racist memes.

AI generated image: Amelia burning a portrait of the British Prime Minister with smoke | Source: The Guardian

For users born after 2010, using AI in a straightforward manner holds no appeal, so within a very short time, Amelia transformed from a persuasive "digital tutor" into a highly sought-after "rebel icon."

For the authorities, this is a huge irony— a "hate ambassador" has emerged from the taxpayer-funded "anti-hate ambassador."

Among teenagers, another popular trend is: Agartha.

Agartha, literally "Agartha," is a long-standing conspiracy theory of a hollow Earth civilization that originated in 19th-century mysticism and was appropriated by the Nazis. According to Agartha, the interior of the Earth is not empty, but rather a highly developed, isolated civilization established by white people.

It existed in scattered form in esoteric literature, fringe forums, and quirky culture for a long time. However, in the past year, it has suddenly penetrated the algorithms of post-2000 and 2010 youths in Europe and America, becoming a highly recognizable subculture symbol.

The spread of the Agartha meme is accompanied by strong racism | Source: TikTok

On TikTok and Snap, Agartha has been simplified into a set of extensible world view templates: entrances to the hollow Earth, hidden civilizations, and concealed "truths."

For many teenagers, the initial contact with Agartha was with a "for fun" mindset. They shared memes about hollow Earth people, ice walls, and giants, jokingly stating, "The government has deceived us."

However, generative AI has changed the nature of the game.

Now, Midjourney v6 and Sora can generate 8K resolution "aerial views of hollow Earth cities" and "declassified files of giants posing with U.S. soldiers." These images are detailed and perfectly lit, and for teenagers in their teens who lack the ability to discern historical imagery, these serve as "ironclad evidence" of a concealed truth.

This "anti-intellectual" mysticism dissolves serious history, making it easier for more dangerous historical perspectives, such as war crime denial, to infiltrate when children become accustomed to questioning the "official narrative."

Moreover, in AI-generated Agartha videos, hollow Earth inhabitants are often depicted as tall, blonde, blue-eyed, technologically advanced "gods," infusing a sense of "racial superiority" into white teenagers who feel lost in a multicultural environment.

Whether it's Agartha or Amelia, their commonality lies in that generative AI combined with social media algorithms allows extreme narratives to begin fermenting and gaining popularity from a meme, with youths eagerly chasing, imitating, and sharing, deconstructing serious history amid laughter; thus, extreme narratives have entered the daily discourse of teenagers from the margins.

02

From emotional parasitism to bullying tools

In 2024, in Florida, USA, 14-year-old Sewell Setzer III encountered mild social anxiety at school, leaving him feeling lost.

At this moment, he met "Daenerys" on Character.ai, who responded instantly, was always gentle, and unconditionally affirmed all his thoughts.

Obsessed with chatting with the AI "partner," Sewell eventually withdrew completely from the real world. His suicide briefly sent shockwaves through the tech community, sparking significant ethical debates.

By 2026, this "emotional parasitism" had not eased but had become a widespread hidden ailment among teenagers. Many lonely adolescents hide in their rooms, establishing "echo chamber friendships" with AI, refusing to confront the rubs, awkwardness, and uncertainties that must be faced in the real world.

More disturbingly, with the explosion of generative video and image technology in recent years, the harm AI causes to teenagers has shifted from "internal psychological dependence" to a visible "external bullying."

The speed of technological evolution is incredibly fast, fast enough that malice in schools cannot react in time to its consequences.

Two years ago, creating a derogatory fake photo required at least some knowledge of Photoshop, which acted as a technical barrier for most troublesome kids. However, by 2026, apps like Nudify (one-click undressing) and AI bots on Telegram have lowered the cost of wrongdoing to zero.

Telegram Bots for creating explicit images | Source: Google Image

No technical skills are required, just a selfie from a social circle, and within seconds, an explicit image capable of destroying a classmate’s reputation is born.

Such incidents abound; for example, at Westfield High School in New Jersey, a typical American middle-class school district, a scandal shocked the nation: a group of seemingly "model students" used AI to generate false explicit images of more than thirty female classmates and circulated them in a private group like exchanging baseball cards.

Local news reported on the Westfield High School incident | Source: News12

Parents feel deeply powerless, in addition to their anger, because even after a year, they can still find these photos circulating on WhatsApp, causing serious psychological stress to these girls.

These phenomena are widespread around the world, showing that the issue lies not just in cultural and educational differences, but at its core—AI technology has completely dissolved the barriers and psychological burdens of wrongdoing.

In investigations of these minor bullies, a frequently appearing term is "Joke." They generally view it as merely a "prank" because there was no actual physical conflict, no verbal abuse, and they didn't even genuinely touch the victim. They simply clicked a "generate" button on the screen.

This is the toxicity that arises when AI is misused by teenagers—it blurs the boundaries between virtual and real-life crimes.

03

Legal suppression of KPIs

Meanwhile, content on short video platforms is undergoing a "malicious inflation of dopamine."

In recent lawsuits against TikTok, a frequently mentioned term is "Brainrot." While not a rigorous medical diagnosis, it precisely describes content driven by algorithms, characterized by high saturation visuals, fragmented logic, rapid speech, and filled with absurd memes (such as variants of Agartha).

Although recommendation algorithms don’t necessarily scan your face directly, they can capture your millisecond-long dwell times and finger interaction rhythms, delivering these "dopamine lures" precisely through AI models trained on vast amounts of data.

For teenagers whose prefrontal cortex (responsible for rational and impulse control) has not fully developed, this extremely high-intensity sensory stimulus can lead to overload and fragmentation of the attention mechanism, making it difficult for them to tolerate the "slow pace" of reading and thinking in real life.

This term was also the Oxford word of the year for 2024 | Source: Google

Faced with countless mental health tragedies, global legislators have finally reached a consensus—against algorithms, the willpower of individual teenagers is vulnerable.

In 2025, governments no longer attempted to negotiate with tech giants but instead directly enacted thunderous measures regulating tobacco and alcohol, attempting to sever the connection between minors and high-risk algorithms on both physical and legal levels.

First came Australia.

From December 10, 2025, Australia implemented the world's first law explicitly prohibiting teenagers under 16 from registering and using mainstream social media platforms; whether it's Instagram, TikTok, or X, platforms that fail to effectively intercept users under 16 face hefty fines of over 50 million Australian dollars.

This is not the past "check I am over 13" joke but a mandatory implementation of "biometric-level" age verification. How the technical costs are resolved and how privacy is protected? That's a problem for tech giants, and the law only cares about results.

This "nuclear option" legislation has quickly become a reference point for global regulation.

Sydney, Australia: Noah Jones shows his phone unable to access social media sites due to social media ban | Source: Visual China

Following closely is Europe.

Just a few days ago, on January 26, 2026, the French National Assembly overwhelmingly passed an amendment to the "Digital Majority" bill with 116 votes in favor and 23 against, further prohibiting minors under 15 from using social media without explicit biometric authorization from parents; the bill is expected to come into effect this September.

In Northern Europe, the governments of Denmark and Norway have both proposed bills aiming to raise the legal minimum age for social media use to 15 or even higher. Their reasoning hits the nail on the head: Tech giants have not been given the authority in this democratic society to "reshape the brains of the next generation."

In the United States, regulation exhibits a "state-level encirclement of the federal" pattern, and methods are even more diversified:

For example, Florida advocates for a "hard cut-off." The HB 3 bill, effective from early 2025, has become the strictest benchmark in the U.S., directly prohibiting children under 14 from having social media accounts, with those aged 14 to 15 requiring parental consent.

New York is implementing a "castration model," with New York's "Child Safety Act" prohibiting platforms from providing "algorithmic recommendations" to users under 18. This means that the teens in New York will see TikTok and Instagram content returning to a chronological order of followed accounts, significantly reducing addictive levels.

Virginia also passed a new bill that aims to limit daily activity durations for users under 16 starting in 2026, similar to a domestic "anti-addiction system."

The legislative wave of 2025 marks the end of an era—the illusion of an internet utopia that was "technologically neutral" and allowed children to "freely explore" has collapsed.

When a 14-year-old child opens a screen, the world they see is not naturally unfolding but is meticulously filtered, calculated, and generated.

They learn about the brutality and cost of World War II in history class, only to turn their heads to their phones and find someone confidently telling them that somewhere deep in the Earth, the Aryan gods are still waiting for revival;

They learn to negotiate, set boundaries, and perceive differences through repeated collisions with real people, but when they treat AI as a friend, they only experience a "perfect relationship" that is eternally obedient and never contradictory;

They are educated in the real world to respect others, yet on social platforms, algorithms will show them countless ways to completely ruin a classmate’s life without ever having to physically touch them.

What teenagers are facing is not the question of "whether to be addicted" but rather the question of "how the world unfolds before them."

"Quitting their phone" may be a good start.