How Social Media and Chatbot Bans Could Backfire

Amanda Third, PhD, has received research funding from the Tech Coalition Safe Online Research Fund and is a member of a number of advisory boards, including the Google Kids and Families Expert Advisory Board and Snapchat’s Safety Advisory Board.

In 2025, Australia passed legislation banning youth under 16 from social media. Since then, France, Indonesia, Malaysia, and some US states have enacted similar restrictions. The United Kingdom and Canada are proposing similar legislation, with Canada’s Bill C-34 also looking to reign in AI chatbot use.

Amanda Third, PhD—codirector of the Young and Resilient Research Centre at Western Sydney University—says that 42 countries are now considering similar restrictions. The laws tap into anxiety over youth social media use—brought to the forefront by Jonathan Haidt’s book, The Anxious Generation, which some experts have criticized for its lack of scientific accuracy—as well as distrust over big tech companies.

Social scientists warn that these bans don’t work and that they may drive youth to less regulated platforms to evade age restrictions. In Australia, “the first compliance report showed that 70% of underage users had remained on social media,” says Third (a peer-reviewed study of over 400 adolescents found the number was closer to 85%). “It seems to me quite negligent to put in place a policy that potentially makes children more unsafe and protects very few.”

The Evidence Backing Restrictions

While there’s a small correlation between the amount of social media use and mental health, Third says the idea that the link is causal is “dead and buried.” She notes that mental health “doesn’t have single isolatable drivers.” Factors like socioeconomic status, the strength of connection with family and community, and exposure to violence have a larger impact. The type of impact can also depend on other factors, like the causes and patterns of use.

Social media restrictions could cut off isolated and marginalized individuals who benefit from social media. “The young people who could probably least afford to be cut off from their communities are the ones who tend to be suffering and getting harmed the most by these types of laws,” says David Greene, JD—senior counsel for the Electronic Frontier Foundation.

Similar concerns were echoed in a recent UNICEF (United Nations Children’s Fund) statement, which noted that social media could be a “lifeline providing access to learning, connection, play, and self-expression.”

‎

The restrictions may also make youth less informed about current events. In Australia, one report found that 51% of teens reported accessing less news directly as a result of the new bans. “These bans are really designed to deny young people their legal right to access information as well as to deny them the opportunity to speak to each other and to the world,” said Greene.

The data on AI chatbots aren’t in yet, but Third expects them to fall within the same lines. Evidence so far suggests that many children are using chatbots as a search engine or for homework help but that chatbot use is also linked to cases of bullying, harassment, grooming, and the production of child-sexual abuse material.

“Swiss Cheese” Legislation

Experts are also raising concerns that the bills imposing bans and restrictions are being rushed through. The Australian bill wasn’t the result of grassroots organization and pressure from parents but, Third opines, was pushed through by politicians who feared having their political campaign derailed by a major Australian-owned news corporation ahead of an election.

The bill was passed in the wake of a 2024 skirmish between the Australian government and social media companies, with News Corp having started the Let Them Be Kids campaign, arguing for an under-16 social media ban.

“We were on the runway to a federal election,” opines Third. No politicians wanted to go against it because “News Corp is notorious for being able to sink particular candidates.”

The result is what Third calls “Swiss cheese” legislation, which fundamentally fails in preventing youth from accessing social media, as they can still view content without logging in. And although social media companies have a problematic track record with protecting youth from harmful content, they still have some safety measures: “What our legislation has done is in fact to remove a key layer of protection for children and young people.”

Canada’s Bill C-34 is also flawed, according to experts. While he thinks some of the proposed regulations for chatbots are reasonable, Michael Geist, LLM, JSD—Canada Research Chair in Internet and E-Commerce Law and professor at University of Ottawa—says that the bill is analogous to an “illusion,” as it’s being pushed through without clarifying important details, including enforcement mechanisms.

“The fact that the government seems to think that it can just sideline some of this to try to march ahead without getting everything it needs in place…is very problematic as well,” says Geist.

A digital safety commission, which would help formulate decisions about enforcement and exemptions, won’t be set up until 18 months after the bill passes. Unlike the legislation in Australia, the commission could exempt certain platforms once they meet adequate criteria showing that they are safe, but the gap in enforcement, say Third and Geist, is extremely problematic.

“I don’t think a court is going to hold a company liable for failing to meet a standard that didn’t exist at the time,” says Geist. The federal government proposing the bill has not consulted with the privacy commissioner either, he adds, which is standard for any bills affecting privacy.

Greene and other experts have raised concerns that these bans also mean that tech companies will need to collect more information from users to determine who is old enough to use these services. In many cases, the legislation will require everyone—not just under 16s—to submit sensitive biometric data for age verification.

“We’re already seeing data breaches among the companies that have the age-estimation technology,” said Greene.

Evaluating the Legislation

Third sits on the advisory panel of a longitudinal Stanford Social Media Lab study set to evaluate the impact of these restrictions.

“The team there is very clear that we are very unlikely in the two years that we are running the evaluation to see the ban have any significant effect on the mental health of children and young people in Australia,” she says. They are, however, expecting that if there is a positive change, it would be in the perceptions among parents and youth about when it’s appropriate to use certain devices and platforms.

Greene also warns parents that these restrictions and bans might further empower governments to determine what type of information is harmful to young people, which “will lead to the state denying young people access to material about their own sexuality, about religion, about their own health.”

While countries continue rolling out bans and restrictions, growing evidence suggests these may be ineffective and may potentially harm those they aim to protect. Regulators will need to work with researchers to understand the scope of potential risk, regulate tech companies without impinging on rights or access to information, and work toward building holistic policy to support youth.