A few months ago, I ran a small experiment. Nothing huge – just an observation. I posted the same text about a person needing help in two different places. First, in a group chat with people I know personally. Second, in a small, niche community where no one knows me. Guess where the reaction was more alive, genuine, and immediate? That's right: in the community of strangers. People who had never met me wrote detailed responses, offered concrete solutions, and expressed support. My acquaintances? A few likes and silence.
I'm not an exception. I've studied this pattern long enough to understand: it repeats itself time and again. And it calls into question everything we think we know about empathy in the digital age.
Before we dive into what's happening to empathy on the internet, we should agree on our terms. Empathy is not just “sympathy.” It's the ability to perceive another person's emotional state as your own, to model their experience within yourself. Neuroscientists talk about mirror neurons – brain cells that activate both when you perform an action and when you watch someone else perform it. Simply put, when you see someone in pain, your brain literally “hurts” a little, too.
This system has been evolutionarily honed to work in person: face, voice, scent, posture, micro-expressions – all are signals our brain reacts to instantly and effortlessly. This is why being there in person is so powerful. It's why you can be indifferent to a statistic – “thousands affected by the flood” – and be completely devastated by a single face in a photograph.
Now, imagine you remove everything physical from this equation. No face. No voice. No scent. All you have is text, a profile icon, and a generic “avatar.” What happens to empathy then?
The answer turns out to be surprisingly complex. The internet doesn't kill empathy – it distorts it. Sometimes amplifying it, other times making it vanish. And what's interesting is that both processes can occur within the same person simultaneously.
On one hand, we see the phenomenon of “viral compassion.” The story of a stranger – someone who lost their job, a critically ill child, a dog rescued from the street – spreads across the web and gathers thousands of sincere reactions. People donate money, write words of support, and share posts to help. Crowdfunding campaigns for strangers regularly raise sums that would otherwise seem impossible. This is real compassion – not an imitation.
On the other hand, these very same people are capable of writing cruel, derogatory things in the comments under that same story to strangers whose opinions differ from their own. The effect of depersonalization – when the person behind the screen is no longer seen as a complete, living individual – is well-documented in social psychology. When you lack the physical presence of the person you're talking to, you also lack the automatic “brake” that engages in a face-to-face conversation.
This is the paradox: the same channel that allows us to empathize with total strangers on the other side of the planet simultaneously lowers the barrier to cruelty toward people we can't see in person.
Let's go back to my experiment. Why did strangers react more vividly than people who know me? It's counterintuitive. But there's an explanation, and it's quite simple.
In real social networks – and I mean real-life relationships, not apps – there are a host of filters. Status. Expectations. A reluctance to “get involved in someone else's business.” A fear of appearing intrusive. People who know you personally carry the burden of context: they remember past situations, they weigh the appropriateness of a response, they think about what the rest of the group will say. The result is silence, born not of indifference, but of social caution.
A stranger on the internet is free from this burden. They have no context, no backstory, no expectations. They see a signal – and they react to the signal. This is precisely what makes online communities such surprisingly supportive spaces at times – especially for people who find it difficult to talk about their problems with those closest to them.
Psychologists call this the “stranger on a train effect”: people often find it easier to open up to a complete stranger they'll never see again than to a friend or relative. The internet has scaled this mechanism up to a global level.
This is where it gets really interesting. Because the processes described above don't happen in a vacuum. They happen on platforms designed to maximize engagement – platforms that have a very clear understanding of what kind of content gets a reaction.
Emotional content drives more engagement than neutral content. This is a long-known fact. But here's the crucial part: algorithms don't distinguish between types of emotion. Anger, fear, affection, outrage, compassion – from an algorithm's perspective, it's all just “high engagement.” And that means it's all promoted with the same aggression.
The result is predictable: our online feeds become a continuous stream of emotional triggers. Stories that bring us to tears. Situations that spark outrage. Videos that make us go “aww.” Over and over, without a pause.
What's wrong with that, you might ask? More emotional stimuli means more empathy, right? No. The opposite happens. Psychologists describe this phenomenon as “compassion fatigue” – a state in which a person, overloaded by others' pain, gradually loses the ability to respond to it. This isn't callousness – it's a defense mechanism of the nervous system. You simply can't care about everyone, all the time, indefinitely.
Compassion fatigue used to be an occupational hazard for doctors, social workers, and journalists covering crises. Now, it's available to anyone with a smartphone and a social media account.
For a long time, I've tried to systematize what I've been observing – in studies, in conversations, and in my own behavior. I've ended up with three main patterns of how empathy gets distorted online.
1. Empathy as Performance
Publicly expressing sympathy online inevitably takes on the features of a performance. When you write “I'm so sorry” under someone's post, are you truly sorry? Probably, yes. But you are also demonstrating your sympathy to an audience. This isn't necessarily a bad thing – public norms of sympathy are important. But the problem is that over time, the form begins to replace the substance. A share becomes “I've done my part.” A broken-heart reaction means “I've expressed solidarity.” If your account is filled with the right markers of sympathy, you must be a good person.
Social psychologist Daniel Batson distinguished between empathic concern (a genuine desire to alleviate another's suffering) and personal distress (the discomfort from observing another's pain, which motivates you to relieve your own discomfort, not to help the other person). Online, the line between these two states is especially blurry.
2. Selective Empathy and the 'In-Group'
It's well known that algorithms create bubbles. But a less-discussed consequence of these bubbles is what they do to our empathy. We begin to empathize primarily with those who are like us: people with similar views, similar lifestyles, similar backgrounds. This is natural in itself – empathy has always worked better “within the group.” But the online environment amplifies this effect to an extreme.
A person from our community who is going through a hard time triggers a wave of support. A person with different views or from another “camp” is more likely to be met with schadenfreude or indifference. This isn't a pathology – it's the predictable outcome of systems that optimize our interactions with people who are “similar.” But on a societal scale, it means our collective empathy is fragmenting into isolated islands, separated by a cold ocean of mutual misunderstanding.
3. Empathy Without Action
This third distortion is perhaps the most insidious. The internet has created the illusion that feeling is the same as doing. You read something, felt moved, left a reaction, maybe even shared it. The emotional cycle is complete. Your brain gets its satisfaction from a “duty fulfilled” – and moves on to the next post.
This phenomenon is sometimes called “slacktivism” – from the words “slack” and “activism.” It's a minimal effort that creates the feeling of participation. The problem isn't that slacktivism is completely useless; sometimes a share really does help. The problem is that it can crowd out more tangible forms of participation – financial support, volunteering, or simply being physically present. The emotion is discharged online – and never makes it to the real world.
If we look at the bigger picture, the data is moderately concerning. Studies conducted over the years at universities in the US and Europe have recorded a decline in empathy scores among young people since the early 2000s – right around the time social media became a mass phenomenon. It's hard to establish a causal link; maybe the internet isn't the culprit, but rather changes in family structures, education, or urban life. But the correlation is stable and replicated across different samples.
At the same time, we're seeing a curious phenomenon: people are getting worse at recognizing emotions from facial expressions. This skill, like a muscle, requires practice – and that means real-life interaction. The more our communication shifts to text, the less we “train” this ability. A study at UCLA found that children who spent five days at a camp without screens were significantly better at recognizing nonverbal cues than their peers with normal levels of screen time.
It's not a catastrophe. But it is a signal.
I can't fail to mention one more aspect that is usually left out of the conversation about “compassion fatigue” and “bubbles.” Empathy is a powerful tool for manipulation. And on the internet, it's used systematically.
Viral stories that trigger a wave of sympathy aren't always real. Sometimes they are fabricated for monetization or to advance a specific agenda. An emotional response is the fastest way to shut down critical thinking. While you're crying over a story, you're not checking the facts. While you're outraged by an injustice, you're not asking questions about the source.
I've fallen for this myself a few times. I'd later find confirmation that the story was inaccurate, embellished, or an outright fabrication – but the emotional reaction had already happened, had already been posted, had already been put to work for someone else. It's an unpleasant feeling. But it's what forces me to keep a simple rule in mind: the harder a story hits you emotionally on the first read, the more thoroughly you need to vet it.
The good news is that empathy isn't a fixed personality trait. It's a skill, and like any skill, it can be developed and can degrade. The bad news is that developing it requires doing things that are increasingly inconvenient in the digital age – meeting people in person, reading fiction, and slowing down.
Here are a few things that, based on the available data, actually work:
- Slowing down before reacting. Before you hit react or write a comment, take a one-second pause. “What am I feeling right now – is this genuine empathy or an automatic reflex?”
- Shifting from passive compassion to concrete action. Instead of just sharing, consider: “I donated five quid,” “I messaged them privately,” or “I showed up to help.” A small action is worth more than a thousand shares.
- Intentionally stepping out of your bubble. Read people you disagree with – not to argue, but to understand what the world looks like from their side. It's uncomfortable. And that's precisely why it works.
- In-person communication as a form of hygiene. Not because “screens are bad,” but because your brain needs real faces, voices, and nonverbal signals to keep its empathy muscles from atrophying.
- Skepticism toward emotional stories. Check the source. Not because everyone is lying, but because the ones who are, are counting on you not to check.
I started this article with a small experiment – and I want to end it with a confession. When I first began to think systematically about this topic, I expected to arrive at a simple conclusion: “The internet is killing empathy, and everything is awful.” It would have been a beautifully pessimistic thesis. I would have framed it nicely.
But it turned out to be more complicated. The internet doesn't kill empathy – it creates a new environment for it, with new opportunities and new traps. The opportunities are real: millions of people receive support from strangers they never would have found in their physical environment. The traps are also real: compassion fatigue, performance over feeling, in-group/out-group selectivity, and manipulation through emotional content.
The difference between a person who emerges from this environment with their empathy intact and one who emerges with it atrophied isn't about how much time they spend online. It's about how consciously they do it.
Your brain adapts to the environment you give it. Create the right environment, and your empathy will be fine. Create the wrong one, and you'll one day find yourself looking at someone else's pain with the same expression you use to check the weather forecast.
The choice, as usual, is yours. Just make it consciously, not automatically.
