Managing Editor
There’s an extremely important difference between doing something because you enjoy it and doing it because you’re addicted. Yet, it can be hard to tell the difference looking in from the outside, or indeed, even for the person who’s potentially slipping into addiction. All too often, that assessment only gets made in retrospect, based on harm suffered as a result of the addiction.
I’d argue that how to make that distinction is one of the most important questions in human behavior facing the world today. It’s a question that spans the fields of business, science, and law. And we’re still far from having a clear answer.
A trial is currently underway in Los Angeles to decide the question of whether Meta deliberately seeks to addict its users. The plaintiff in this particular case is a young woman who says she experienced severe anxiety, body dysmorphia, and online bullying as a result of social media addiction. However, it’s just one of thousands of similar cases being brought against various online entertainment companies, including social media, video games, and, of course, gambling and “gambling-adjacent” products.
What’s noteworthy about the Meta case is that it has reached the trial stage. Most similar cases end in settlements without an admission of wrongdoing, so that the defendants are able to negotiate what adaptations they’ll make to address the concerns. Indeed, Snapchat and TikTok were also defendants in this case, but settled early. Meta’s determination to see things through could set an important precedent one way or another.
Sportico, in covering the case, argues that the result will prove to be important to sportsbooks and prediction markets. They’re not wrong, but such cases will never truly be a matter of settled law until there’s some clarity on the addiction vs. engagement issue.
Keep Reading
The Consequences of Metric-Driven Product Development
It’s easy, and often justifiable, to vilify the companies that produce addictive products. Even so, it’s instructive as a starting point for thinking about this to look at how addictive properties can emerge from even well-intended product development.
Part of my job is to improve PokerScout’s news product. The way I do that is largely by tinkering with it and seeing the results. My main metric is the number of engaged seconds that readers collectively spend on each article. If I try something and engagement goes down, I eventually abandon that experiment. If it goes up, I instruct my team to write more news on that topic, or in that format, or using that approach to framing.
Maybe you’re going to say that’s fine because news isn’t an addictive product. And to my knowledge, no one is flunking out of college or missing their daughter’s wedding because they can’t tear themselves away from PokerScout. But “to my knowledge” is a key phrase, there. If I were causing harm in my efforts to improve the product, I might not be aware of it. Moreover, I actually do think some of the more sensationalist or politically polarizing forms of journalism could justifiably be classified as addictive.
Perhaps addictiveness isn’t a binary property. It’s probably a spectrum, with heroin at one end and table lamps at the other. Gambling, video games, social media, and, indeed, news all fall somewhere in between. The degree of care that should naturally be required in product development depends on where you fall. But, especially when it comes to new product classes, you might not know where that is right away.
So, how can you tell if your higher engagement metrics are actually addiction indicators?
What Psychiatrists and Psychologists Say
Differentiating engagement from addiction is a hot topic in psychology and has been for over two decades, particularly when it comes to video games.
Gambling addiction — under one name or another — has been in the Diagnostic and Statistical Manual of Mental Disorders (DSM) since its third edition in 1980. At the time, it was called “pathological gambling,” though the current term is “gambling disorder,” and it’s classified alongside substance addictions. By contrast, “internet gaming disorder” is still in the proposal stage and identified as a topic for further research, alongside things like caffeine addiction that haven’t yet received official status.
Much of the research on the topic is coming out of Asia, where public concern around the problem is even higher than it is in the West.
Unfortunately, the only thing the psychological community agrees on is that when it comes to video games, differentiating addiction from engagement in a hobby is going to be important. Some psychiatrists want to move more quickly on diagnosis and treatment than others. One 2025 study out of Japan summarizes the disagreement as follows:
[Some researchers] argued that the premature application of a clinical diagnosis would cause moral panic, lead research to be locked into confirmatory thinking rather than exploratory thinking, and stigmatize the healthy majority of gamers. [Others] emphasize that clinical and public health needs to recognize the danger of excessive and problematic gaming.
That study concludes that most proposed indicators can’t differentiate engagement from addiction. It recommends that the focus should be on loss of control as the key feature separating the two.
Unfortunately, it isn’t possible to take someone at their word when it comes to that. There’s a reason the phrase “I can stop anytime” is stereotypically associated with gambling sufferers.
So, we’ve merely replaced “how do we tell engagement from addiction?” with a related question: How can we tell if someone’s out of control?
The Problem With Focusing on Harm
Related to loss of control, psychiatric definitions of addiction often include the concepts of “precedence” and “continuance.”
Precedence means prioritizing the subject of the addiction ahead of things like family, friends, work, and health. Continuance means persisting in engaging with the thing in question even after suffering harm on those fronts.
On the one hand, that seems like a fairly clear-cut way of differentiating addiction from engagement. If you’re suffering harm from what you’re doing but still can’t stop, then you’ve got an addiction.
On the other hand, the purpose of trying to make this distinction in the first place was to prevent harm. If we can only decide a product is potentially addictive after it has created a crisis, we will forever be left trying to close the barn door after the horse has bolted.
Unfortunately, the law also has a tendency, at times, to be results-oriented. Two similarly reckless drivers who run the same light might be treated very differently depending on whether or not another car happened to be coming along the crossing street as they were doing it.
That’s especially true when it comes to civil law, where it’s not even possible to sue based on hypothetical harm. So, of two product developers who were both chasing engagement, only the one who happened to produce an addictive product will get in trouble. And, more importantly, that will only happen after the harm has already been done.
The Answer May Lie in Usage Distribution
I don’t think this is an unresolvable problem. However, I think addressing it requires a shift in perspective away from individuals and toward collective behavior.
Another engagement-addiction study from 2015 looked at users of a particular multiplayer online role-playing game. The researchers noted that although many positively-engaged players committed significant hours to the game, their average play time per week was 12 hours, compared to 40 hours for those who were showing signs of addictive harm.
That doesn’t mean that we can specify some cutoff and say that pursuing engagement beyond that limit counts as developing an addictive product. That would be overly simplistic. However, I do think that the ultimate answer lies in the intensity of use across a sample group.
Earlier on, I hypothesized a spectrum of addictiveness from table lamps to heroin. As a thought experiment, how do you think the distribution of table lamp usage looks across the American population? What about heroin? Or all the things in between, like social media, video games, and gambling?
Classical economist Vilfredo Pareto coined the 80/20 principle, that for most products, 20% of consumers will account for 80% of consumption. If that’s normal, then what does that say about a business for which the number is 10%, or 5%, or 1%? What is that number for Facebook? For the average casino game?
In short, pursuing median engagement may be good business practice. But driving up mean engagement can lead to harmful outcomes if the gains are coming predominantly from users who were already highly engaged, thereby skewing the usage distribution toward the high end.
That may be the lens that the legal and scientific worlds need to use. Unfortunately, it’s already in heavy use in the business world, where targeting “whales” is still seen as good practice.
