The details on "Project Mercury" are particularly striking when you look at the transcripts from Ottawa. During the House Heritage Committee hearings for the Online Harms Act (Bill C-63), tech representatives repeatedly told MPs that the science linking platforms to mental health issues was inconclusive. We now know their internal data confirmed the link years ago. It is frustrating to see the "correlation isn't causation" defense used to stall Canadian legislation when their own researchers had already settled the debate.
Canada will 100% use this to censor dissent instead of protect children. That’s why even when I know all the damage of social media, I’m still against any regulation
We don’t need to wait another ten years of playing it safe while kids pay the price right now. This wasn’t some careful test; it was a huge, uncontrolled experiment on kids, and the signs are clear. Anxiety, depression, bullying, exploitation, and attention issues didn’t just pop up out of nowhere. They’re linked directly to smartphones and social media becoming a normal part of childhood.
Waiting for perfect agreement is a luxury parents and governments can’t afford. When the danger is obvious, responsible leaders step up. Australia raising the minimum age isn’t overreacting; it’s basic risk management.
We already see improvements when phones are taken out of classrooms. Extending that to early teens makes sense. Less social media means fewer predators, less humiliation, and more space for kids to live and learn in the real world.
The experiment failed. It’s time to end it instead of pretending to be neutral while another generation suffers the consequences.
Well said. Even though I was never a parent, it concerns me greatly that children are being exploited this way, and the harm it’s causing them, up to and including self harm and even death. Apart from this, I have even seen older adults become highly addicted to computer games, hours a day, to the point where normal conversation dies out. They don’t realize that this habit places them at higher risk for cognitive decline, IMHO.
You are right. The harm to children is real and immediate, and it is not limited to online interactions. Addiction to games or social media at any age rewires attention, diminishes real-world engagement, and carries long-term risks for mental and cognitive health. Adults and children alike are paying the price for habits that were never meant to be normal. The more we recognize this, the more urgent it becomes to set boundaries and protect the most vulnerable.
Jon and Zach, this is invaluable work. The seven lines of evidence and Meta's internal research form exactly the kind of case that parents and policymakers need.
I'd add a mechanism underneath the findings: the architecture itself. Social media isn't just correlated with anxiety and depression. It's structurally designed to keep the nervous system in open loops. Every notification opens something that doesn't close. The body accumulates this. Not sleep debt. Completion debt. Rest doesn't resolve it because the issue isn't depletion. It's unfinished time.
This helps explain why reduction experiments show benefits. It's not just less exposure to harmful content. It's giving the nervous system permission to close loops, to settle, to get an all-clear that the platform is designed to withhold. Completion ends the session. That's bad for engagement metrics.
Your point about "industrial scale" harm is exactly right. And the mechanism scales too. Keep a generation in open loops and you don't just produce anxiety. You weaken the capacity for sustained attention, for memory, for staying with anything long enough for it to become consequence.
The product safety question and the temporal design question are the same question.
I wrote about this last week in "You Are Not Distracted. You Are Unfinished." Thank you
Attributions of accountability are urgently required. Zuckerberg et al have made literally billions of dollars largely off the back of causing a significant amount of mental health problems, and that cannot be allowed to stand. I have no problem with people making a lot of money from their own efforts, but when that money comes with and/or causes huge societal problems, that is something that society shouldn't and mustn't tolerate. Western governments (because only they can) must investigate these issues immediately and urgently, and mete out appropriate penalties.
Compelling need to protect our children is very clear, but the negative toll it is taking on adults is not to be ignored. As I believe I heard you say that the Nantucket Project a couple of years ago, Meta may be a challenge to our democracy that it will struggle to overcome!
So, given the leaks and timelines, did Zuk commit perjury before Congress? What about the company reps who testified to Canada? Can anyone be brought up on charges for their lies? I bet not.
If it can be proven that they lied to Congress or to a parliamentary committee, they should be brought up on charges. Proving that they lied knowingly will be the hurdle, as usual.
I'm sure Meta won't let this crisis go to waste, as age verification becomes another method to harvest and monetize our most sensitive data: biometrics.
I don't need or want the gov't to step in and be my parent. I agree the evidence is strong that young teens struggle to manage the temptations offered by social media and social media offers predators easier access to teens. I definitely don't think the evidence shows the need to hit the problem with a sledgehammer by requiring age verification to access social media or the internet. Show the evidence to parents, schools, teachers, and education unions.
1. Help them ban cell phone use in class.
2. Help parents push back on the social pressure to get their kids smart phones instead of a simple flip-phone.
3. Help law enforcement identify and track the predatory use and techniques, then make an example out of them.
Why does every problem have to be solved with poorly written laws and gov't money/regulation/enforcement?
Any research forthcoming on the searchable hardcore pornography database of young men’s primal fears about sex and women that we ran as an experiment on our 12 year olds? Probably not.
ain't that odd, the mainstream media and Meta align to strangle the internet. War on free speech by freaks fisting cameras hardwired to a walls sight and sound. They are doomed -- we are never going back to the parasitical propaganda pimps of the past 3,000 years. What a beautiful time to be alive!
As a former journalist who covered this extensively, the causal evidence now reads less like a moral panic and more like the evidentiary record finally catching up. Long overdue.
I assume many on this site care about accuracy. The comparison here between the effects of child maltreatment vs social media on teen mental health is completely wrong. The PAF measure of child abuse effects is approximately 5 times more powerful than the very weak d value used to measure the effects of social media. That is, a PAF directly tying child maltreatment to 22% of poor mental health and 41% of suicide attempts is a vastly bigger effect than a d value of 0.22 for social media, which associates social media use with just 1% to 2% of poor mental health (and zero connecting social media use to teen suicide attempt).
The dismissal of child abuse and parents' addiction, violence, depression, and crime as huge factors in teens' poor mental health and safety is a real problem on these blame-social-media sites. It would be highly informative if those on this site would ask for a comparison of "direct harms" involved in (a) teens' being online, vs (b) teens' being in real-world activities such as school, church, families, sports, etc. I realize there are other values involved, but the implication that going online is unacceptably dangerous while being in the real world is completely safe is an obvious falsehood. I think a solid case can be made that online and offline lives are integrated rather than oppositional, but I would at least like to see honest evaluations.
I'm wondering why I’m listening to Cal Newport on my tablet talking about overlooked tech dangers and mentioning Haidt several times and suddenly this post appears in my Substack feed, restacked by someone I don't follow. Yet they say our phones aren't listening or watching us.
This is not an experimental study, it is a longitudinal observational study. That means anytime the term 'effect' or 'impact' appears, it is not actual effect or impact but one in some model created by the study authors to simulate reality. The obvious question is the accuracy of this particular model in approximating actual causation.
Another question is implementation. Programming errors do occur. The paper does not give a link to the code. We plan to request the code and examine it when we find the time.
As to the data, it will not be released until 2026, and only partially then. That means the study is essentially unverifiable now, and may remain unverifiable forever.
What we do know is that these result appear to clash with the results from the U.S. ABCD data set. So the question is why, and it is not an easy question to answer, certainly not quickly (I myself did not hear about the study until a few days ago).
There are, IMHO, plausible explanations that involve methodologies, but that a complex matter that will need to be addressed in a separate article, not a comments section.
The details on "Project Mercury" are particularly striking when you look at the transcripts from Ottawa. During the House Heritage Committee hearings for the Online Harms Act (Bill C-63), tech representatives repeatedly told MPs that the science linking platforms to mental health issues was inconclusive. We now know their internal data confirmed the link years ago. It is frustrating to see the "correlation isn't causation" defense used to stall Canadian legislation when their own researchers had already settled the debate.
Canada will 100% use this to censor dissent instead of protect children. That’s why even when I know all the damage of social media, I’m still against any regulation
We don’t need to wait another ten years of playing it safe while kids pay the price right now. This wasn’t some careful test; it was a huge, uncontrolled experiment on kids, and the signs are clear. Anxiety, depression, bullying, exploitation, and attention issues didn’t just pop up out of nowhere. They’re linked directly to smartphones and social media becoming a normal part of childhood.
Waiting for perfect agreement is a luxury parents and governments can’t afford. When the danger is obvious, responsible leaders step up. Australia raising the minimum age isn’t overreacting; it’s basic risk management.
We already see improvements when phones are taken out of classrooms. Extending that to early teens makes sense. Less social media means fewer predators, less humiliation, and more space for kids to live and learn in the real world.
The experiment failed. It’s time to end it instead of pretending to be neutral while another generation suffers the consequences.
Well said. Even though I was never a parent, it concerns me greatly that children are being exploited this way, and the harm it’s causing them, up to and including self harm and even death. Apart from this, I have even seen older adults become highly addicted to computer games, hours a day, to the point where normal conversation dies out. They don’t realize that this habit places them at higher risk for cognitive decline, IMHO.
You are right. The harm to children is real and immediate, and it is not limited to online interactions. Addiction to games or social media at any age rewires attention, diminishes real-world engagement, and carries long-term risks for mental and cognitive health. Adults and children alike are paying the price for habits that were never meant to be normal. The more we recognize this, the more urgent it becomes to set boundaries and protect the most vulnerable.
Jon and Zach, this is invaluable work. The seven lines of evidence and Meta's internal research form exactly the kind of case that parents and policymakers need.
I'd add a mechanism underneath the findings: the architecture itself. Social media isn't just correlated with anxiety and depression. It's structurally designed to keep the nervous system in open loops. Every notification opens something that doesn't close. The body accumulates this. Not sleep debt. Completion debt. Rest doesn't resolve it because the issue isn't depletion. It's unfinished time.
This helps explain why reduction experiments show benefits. It's not just less exposure to harmful content. It's giving the nervous system permission to close loops, to settle, to get an all-clear that the platform is designed to withhold. Completion ends the session. That's bad for engagement metrics.
Your point about "industrial scale" harm is exactly right. And the mechanism scales too. Keep a generation in open loops and you don't just produce anxiety. You weaken the capacity for sustained attention, for memory, for staying with anything long enough for it to become consequence.
The product safety question and the temporal design question are the same question.
I wrote about this last week in "You Are Not Distracted. You Are Unfinished." Thank you
https://yauguru.substack.com/p/you-are-not-distracted-you-are-unfinished?r=217mr3
Jean Twenge and you guys should ask the question: has modern internet pornography’s celebration destroyed a generation?
Attributions of accountability are urgently required. Zuckerberg et al have made literally billions of dollars largely off the back of causing a significant amount of mental health problems, and that cannot be allowed to stand. I have no problem with people making a lot of money from their own efforts, but when that money comes with and/or causes huge societal problems, that is something that society shouldn't and mustn't tolerate. Western governments (because only they can) must investigate these issues immediately and urgently, and mete out appropriate penalties.
Compelling need to protect our children is very clear, but the negative toll it is taking on adults is not to be ignored. As I believe I heard you say that the Nantucket Project a couple of years ago, Meta may be a challenge to our democracy that it will struggle to overcome!
Thank you for your tireless work surfacing this information and presenting it in a way that can make a difference for average parents! So grateful!
So, given the leaks and timelines, did Zuk commit perjury before Congress? What about the company reps who testified to Canada? Can anyone be brought up on charges for their lies? I bet not.
If it can be proven that they lied to Congress or to a parliamentary committee, they should be brought up on charges. Proving that they lied knowingly will be the hurdle, as usual.
Yes. Essentially only if they participated in or signed off on the research prior to testimony.
I'm sure Meta won't let this crisis go to waste, as age verification becomes another method to harvest and monetize our most sensitive data: biometrics.
I don't need or want the gov't to step in and be my parent. I agree the evidence is strong that young teens struggle to manage the temptations offered by social media and social media offers predators easier access to teens. I definitely don't think the evidence shows the need to hit the problem with a sledgehammer by requiring age verification to access social media or the internet. Show the evidence to parents, schools, teachers, and education unions.
1. Help them ban cell phone use in class.
2. Help parents push back on the social pressure to get their kids smart phones instead of a simple flip-phone.
3. Help law enforcement identify and track the predatory use and techniques, then make an example out of them.
Why does every problem have to be solved with poorly written laws and gov't money/regulation/enforcement?
Any research forthcoming on the searchable hardcore pornography database of young men’s primal fears about sex and women that we ran as an experiment on our 12 year olds? Probably not.
Why is that?
ain't that odd, the mainstream media and Meta align to strangle the internet. War on free speech by freaks fisting cameras hardwired to a walls sight and sound. They are doomed -- we are never going back to the parasitical propaganda pimps of the past 3,000 years. What a beautiful time to be alive!
As a former journalist who covered this extensively, the causal evidence now reads less like a moral panic and more like the evidentiary record finally catching up. Long overdue.
I assume many on this site care about accuracy. The comparison here between the effects of child maltreatment vs social media on teen mental health is completely wrong. The PAF measure of child abuse effects is approximately 5 times more powerful than the very weak d value used to measure the effects of social media. That is, a PAF directly tying child maltreatment to 22% of poor mental health and 41% of suicide attempts is a vastly bigger effect than a d value of 0.22 for social media, which associates social media use with just 1% to 2% of poor mental health (and zero connecting social media use to teen suicide attempt).
The dismissal of child abuse and parents' addiction, violence, depression, and crime as huge factors in teens' poor mental health and safety is a real problem on these blame-social-media sites. It would be highly informative if those on this site would ask for a comparison of "direct harms" involved in (a) teens' being online, vs (b) teens' being in real-world activities such as school, church, families, sports, etc. I realize there are other values involved, but the implication that going online is unacceptably dangerous while being in the real world is completely safe is an obvious falsehood. I think a solid case can be made that online and offline lives are integrated rather than oppositional, but I would at least like to see honest evaluations.
Well said as usual, Mike! Shout it from the rooftops!
I'm wondering why I’m listening to Cal Newport on my tablet talking about overlooked tech dangers and mentioning Haidt several times and suddenly this post appears in my Substack feed, restacked by someone I don't follow. Yet they say our phones aren't listening or watching us.
And yet this recent study from the University of Manchester claims social media time does not increase teenagers’ mental health problems ( https://www.theguardian.com/media/2026/jan/14/social-media-time-does-not-increase-teenagers-mental-health-problems-study ). How can we debunk this?
This is not an experimental study, it is a longitudinal observational study. That means anytime the term 'effect' or 'impact' appears, it is not actual effect or impact but one in some model created by the study authors to simulate reality. The obvious question is the accuracy of this particular model in approximating actual causation.
Another question is implementation. Programming errors do occur. The paper does not give a link to the code. We plan to request the code and examine it when we find the time.
As to the data, it will not be released until 2026, and only partially then. That means the study is essentially unverifiable now, and may remain unverifiable forever.
What we do know is that these result appear to clash with the results from the U.S. ABCD data set. So the question is why, and it is not an easy question to answer, certainly not quickly (I myself did not hear about the study until a few days ago).
There are, IMHO, plausible explanations that involve methodologies, but that a complex matter that will need to be addressed in a separate article, not a comments section.
Debunk it? Perhaps actually look into it first. You might actually learn something.