Flat Cartoon Vector Illustration.
What kids are experiencing today on social media is unlike anything prior generations have had to contend with. The influence of social media on youth mental health is shaped by many complex factors, including, but not limited to, the number of time children and adolescents spend on platforms, the type of content they consume or are otherwise exposed to, the activities and interactions social media affords and the degree to which it disrupts activities that are essential for health like sleep and physical activity. According to a recent surgeon general’s advisory note:
- Over 95% of teens have access to smartphones, exposing them to a digital world saturated with potentially harmful content.
- Depression and anxiety rates among youth have surged, with experts linking excessive social media usage to increased mental health issues.
- A staggering 70% of teens report cyberbullying experiences, contributing to psychological distress and decreased self-esteem.
A few stats that should also be considered:
- According to Risky by Design, a research project by 5Rights Foundation, 75% of the top social media platforms use AI to recommend children’s profiles to strangers.
- Leaked documents from Instagram said the platform is exacerbating body image problems for one in three teen girls, and it’s easy to understand why if we look at the way algorithms display content
- A report from Common Sense Media discovered that the average screen time for American teens increased by 17% in the last two years to over 8.5 hours a day. That’s a full workday and then some.
We’ve reached a point at which tech companies are no longer capable of denying the negative impact they’re having on our children. This isn’t one study, it’s multiple studies over time. Two weeks ago, United States Surgeon General Vivek Murthy joined in with an official advisory note representing the dangers at hand.
The advisory note detailed that widespread social media use among kids and teens poses a significant mental health risk that needs to be addressed immediately. This is a big deal considering a surgeon general’s advisory note is “reserved for significant public health challenges that require the nation’s immediate awareness and action.” They are delivered to call the American people’s attention to an urgent public health issue as well as recommendations for how it should be addressed.
Even before Murthy’s note we’ve seen a wave of child data protection bills and legislation attempting to resolve this problem. Moreover, legislative interest seems to be picking up because lawmakers are realizing it’s hard to vote against protecting children, and there’s a lot of money in enforcing penalties against the wealthiest companies in the world.
The problem with most of the laws being proposed is that even with the best intentions they’ve mostly increased surveillance and they’re making it a legal requirement to profile children. I’ve been in the room for the drafting of a couple of these bills, and I can safely say that nobody in the room wants to make the situation worse but creating a bill that is 1) meaningful enough to make an impact and 2) is simple enough for other legislators to understand and support is no small task.
What’s Happening And Where
According to HuschBlackwell’s State Children’s Privacy Law Tracker, there are more than 40 bills that are actively being pursued across 22 different states, four of which have been enacted: California’s Age Appropriate Design Code, Utah’s SB 152 and HB 311, Arkansas’ SB396, and, as of just the other day, Louisiana’s HB61. There are two common standards for identifying children: the Actual Knowledge Standard and the Constructed Knowledge Standard, both of which we’ll dig into later in the article. Two of the three that have passed require actual knowledge of the children’s age. In short, this requires more invasive identification and authentication than if the bill weren’t enacted at all.
The Actual Knowledge Standard
Actual knowledge standards refer to the use of concrete, verifiable information to determine a user’s age. This typically involves collecting official documents or verifying personal information through trusted sources. Examples of actual knowledge standards include:
- Government-issued identification: Users may be required to provide a scanned or digital copy of their government-issued identification, such as a driver’s license or passport, to prove their age.
- Credit card verification: Some platforms may use credit card information to verify a user’s age. This method relies on the assumption that individuals under a certain age do not possess their own credit cards.
- Age verification services: Specialized third-party services can be utilized to verify age by cross-referencing personal information with official databases and records.
Actual knowledge standards are considered more reliable as they are based on authoritative and verifiable data sources. However, the standard raises privacy concerns about requiring users, especially child users, to disclose more sensitive information than they already do. Additionally, the standard creates accessibility issues because, until a certain age, kids may not have the appropriate ID or credit card access to verify their identity.
Worse, submitting these docs also means companies are required to process this information (albeit temporarily if they’re doing it right), which adds more stress on their privacy and security programs. We’ve already seen that Facebook can’t be trusted with kids’ data. What makes a rational person believe it’s going to be a great idea to require even more sensitive information to be uploaded?
The Constructed Knowledge Standard
Just as it sounds, the Constructed Knowledge Standard involves the use of indirect methods such as algorithmic inference to estimate a user’s age based on various factors. These methods do not rely on verifiable, concrete evidence but instead, make assumptions based on user-provided information, behavioral patterns or a combination of both. Examples of constructed knowledge standards include:
- Self-declaration: Users are asked to enter their birthdate or indicate their age during the registration process. However, this method is easily manipulable, as users can provide false information. This may then be paired against other measures to validate whether the response was true or not.
- Online quizzes or surveys: Some platforms use quizzes or surveys with questions about pop culture, historical events or age-related topics to estimate a user’s age. However, this method can be inaccurate and easily tricked.
- Behavioral analysis: Analyzing user behavior, such as browsing patterns, interests or interactions, can be used to make assumptions about their age. However, this method is prone to errors and may not accurately reflect the user’s actual age.
Constructed knowledge standards are less reliable than actual knowledge standards, as they rely on assumptions and subjective interpretations. They are often employed when direct verification methods are impractical or not feasible.
The problem with both of these standards is that they make the situation worse. You could claim that the constructed method isn’t as invasive as requiring access to government IDs, facial recognition or other standard measures that require more sensitive information to be given away and properly protected but isn’t it just as bad when a law requires a company to profile children unknowingly? That means they’re legally required to create even more detailed and stringent profiles than they already do. And if that’s a black box process how do we know they’re only using that data for verification?
More But Less Invasive Protections
With more than 200 privacy laws across the world, many of which have popped up in the last five years, it’s clear that privacy is a big concern around the globe. Due to the nature of the digital world we live in, invasion of privacy is no longer just a problem for the poor and middle class, it’s become an issue for the elite as well, so change is happening fast. Privacy bills to protect the general public are failing but privacy protections for children is hard to vote against. And with more than $2 trillion in fines over the last five years, it’s clear that privacy makes a great revenue stream for the government.
The problem with the laws that have been passed or are being proposed is that often bills are either 1) entirely toothless and unenforceable, 2) not in line with the realities of the internet, damaging innovation, or 3) making surveillance even worse and increasing risk. Sometimes all three at once.
To date, the burden of protecting youth has fallen predominantly on children, adolescents and their families. At the core of the problem is the addictions built into these platforms, but we should not be regulating design patterns with the hope it’s going to change things. We have to remember that the intent is not to addict children; it is to aggregate as much data as possible about the world, and children just so happen to be in the crosshairs.
If we regulate interface design, new design patterns will pop up because it’s not about the design, it is about the data. It is data, not addicted children, that drives these companies, which means we need privacy protections online just as much or more than we need design and architecture regulations.
Do Not Track (DNT) Children
If age verification continues to be a problem, which it will, we will need to take a step back and look at the issue from a different angle. For example, instead of trying to guess who’s using a platform (constructive knowledge standard) or requiring extra sensitive information (actual knowledge standard) maybe the law should manifest in a way that is more direct and enforceable yet requires less surveillance. Perhaps this process requires documentation at the purchase of the device itself and the law is directed at the service provider level rather than the browser level where millions of organizations have to be policed.
If parents were required to document that the phone is being purchased for a child, we wouldn’t need anyone to upload a government ID nor would we need to create black-box profiling systems by independent third-parties. The law would need to require companies to look for and recognize this binary signal from the device itself, similar to a Do Not Track (DNT) signal. Forced recognition is critical because companies rarely recognize DNT signals from consumers because there’s no backbone to it.
Ban Smartphones Sales Under Age 18
Another option is that we just ban kids of a certain age from having smartphones altogether. This again would be something that would require a decision at the purchase of a phone. That doesn’t mean kids can’t have feature phones to keep in touch with their families and for moments of public safety. It doesn’t mean kids can’t use tablets. It doesn’t prevent them from logging in on a desktop computer or laptop. It would prevent service providers from selling smartphones to children, therefore preventing children from having an addiction in their pocket. Because let’s be honest: when was the last time we pointed out these addictions while kids are carrying laptops around school glued to social media?
The problem is social media, but the most problematic access point is the smartphone. Requiring kids to access the media through another device would add some friction and likely reduce the scale of the addictive behavior quite rapidly.
I know many readers’ initial thought is going to be about how extreme banning smartphones under age 18 is, but take a moment to consider the realities of the world. Within Gen Z there’s already a movement to get away from smartphones. These kids are more aware than any generation about the fact that these devices are terrible for their mental health, ruining relationships with friends and family, and they’re self-selecting out. Additionally, it’s well documented that parents who work in tech tend to not let their children have access to these devices until much later than other families. There’s clearly a problem if the people building the product don’t give it to their children, no?
If we’re so concerned about kids' mental health, why don’t we just follow the trend and create a bill that bans parents from purchasing smartphones for their kids just like kids can’t have cigarettes? Companies are already developing alternative phones that have maps, music and other valuable features but lack the full smartphone interface. They can sell those instead. Kids need access to a phone and the utilities that come along with it, but they don’t need access to everything a smartphone has to offer. And if there’s a law like this in place it takes the burden off parents because now the government becomes the bad guy, not mom and dad.
Protect Kids, Save Families And Regenerate Mental Well-Being Around The Globe
Nearly 70 percent of parents say parenting is now more difficult than it was 20 years ago, with technology and social media as the top two cited reasons. Having spoken to thousands of families since The Social Dilemma came out, it’s easy to see where that sentiment comes from. But if you can’t understand, consider this: millions of parents from all walks of life want to make better decisions for their kids, but their biggest enemy is the kid next door, whose parents allow the kid to use a smartphone. This puts parents in the position of bad cop. Banning the sale of smartphones removes the blame from parents and places it on the government.
I believe a ban on smartphones for children and adolescents would drive increases in familial well-being across our nations, reduce suicide and depression rates among youth and may even develop a more united country with less fighting amongst family members about why the kids aren’t allowed to have a phone.
What Is Age Appropriate?
Thirteen is a common age fought for because it aligns with COPPA, but we’re seeing COPPA 2.0 extend the age limits as far as 16 years old. The United Kingdom and California’s Age Appropriate Design Code both protect children and adolescents under the age of 18. Utah’s S.B. 152, Arkansas’ SB 396 and Louisiana’s HB61 all require children and minors under 18 years old to obtain parental consent. These age limits are based mostly on legal precedence but a push toward 18 seems to be where things are headed.
What’s not talked about enough in this conversation is that 16 years old is also the age at which most kids start driving. Consider one out of every four car accidents in the United States is caused by texting and driving, and texting while driving is six times more likely to cause an accident than driving drunk. Driving is more dangerous than ever before, and it’s because of our smartphones. The National Safety Council reports that cell phone use while driving leads to 1.6 million crashes each year! Considering these numbers, perhaps 18 is more appropriate so these new drivers can spend a couple of years learning to drive without distractions.
We can argue about what age and we can discuss what tactics need to be implemented, but there’s one thing no one can fight: there’s no better time to protect our children than today, and you can take that advice from the United States surgeon general, not just me. What’s extreme today will be common tomorrow. It’s time to add friction to the surveillance economy now before things get worse.
