When Empathy Becomes Dangerous: From Bias to “Suicidal Empathy”

“The biggest deficit that we have in our society and in the world right now is an empathy deficit.” – Barack Obama

“Arguing against empathy is like arguing against kittens.” – Paul Bloom

Empathy has become something close to a sacred value in modern culture. In therapy, politics, education—everywhere—we’re told to be more empathetic.

And on one level, this makes perfect sense. A world without empathy would be cold, indifferent, and brutal. But can empathy be a problem? It sure can.

Paul Bloom once said that arguing against empathy is like arguing against kittens. Well, that’s what I’m going to do. I think that when something can’t be questioned, it’s a problem. Some ideas become so morally untouchable that we stop examining them.

This blog explores two versions of the argument against empathy: the measured, evidence-based case from Against Empathy by Paul Bloom, and the more provocative argument from Gad Saad in his upcoming book Suicidal Empathy.

Empathy as a Cognitive Bias

Bloom’s central claim is simple: Empathy—defined as feeling what others feel—is a poor guide to moral decision-making. This cuts against a deeply held intuition. We tend to assume that feeling more makes us better people. But Bloom argues the opposite: empathy systematically distorts our judgment. How?

Because empathy is not neutral. It is biased. Empathy functions more like a spotlight than a floodlight. It illuminates certain people intensely while leaving others in the dark.

It zooms in on identifiable individuals (the “one sick child” effect), people who are similar to us and emotionally vivid stories, while it frequently ignores statistical realities, long-term consequences and people who are less visible or relatable.

There is also evidence that empathy is not evenly distributed. Psychologists (e.g., see studies by Emile Bruneau and colleagues) have long shown that our capacity for empathy is “parochial”—it is stronger for those we identify with. And when empathy is directed intensely toward one group, it often comes at the expense of another group. Studies suggest that feeling empathy for one side in a conflict can reduce concern for the other, and in some cases even increase hostility toward them. The way this looks in the real world is people who often talk the most about being 'empathic' or 'compassionate' are often no more empathic or compassionate than anyone else, they are just more selective or tribal. Actually that's being too generous. In one clever study, it was found that empathy can also make us actively spiteful toward others if we feel they’ve troubled the object of our empathy and that those highest in empathy (those more sensitive to vasopressin and oxytocin) are typically more spiteful. The clearest example of this today might be the way Palestine supporters often think about Jews. But we might also think about how people who claim to care about the poor talk about the rich.

Bloom’s alternative is what he calls rational compassion: care about others but don’t rely on emotional resonance. Instead, use reasoning, evidence, and proportionality.

Gad Saad takes this further and uses stronger language. Using the term “suicidal empathy”, he describes many situations where empathy overrides basic self-preservation—at the level of individuals, groups, or entire societies.

From an evolutionary perspective, empathy evolved in small, tight-knit groups (eg. kin, tribe, reciprocal relationships). In those environments, empathy generally aligned with survival.

But in modern, large-scale societies, that empathy can be extended too broadly resulting in the prioritising offenders over victims, extending care in ways that undermine stability, and valuing emotional expression over long-term outcomes. Whether or not one agrees with his specific examples (and many won’t), the underlying claim is worth taking seriously: A virtue, when unconstrained, can become a liability.

From a clinical perspective, this tension between empathy and reason isn’t just philosophical, it’s something I work with every day. In Cognitive Behavioural Therapy (CBT), one of the core aims is to help people step back from their immediate emotional reactions and examine them more carefully. Clients learn that feelings, while powerful, are not always reliable indicators of truth. A surge of anxiety doesn’t necessarily mean danger is present. A sense of guilt doesn’t always mean wrongdoing has occurred. In this sense, CBT is quietly radical: it teaches people not to automatically trust their internal experiences, but to evaluate them. To ask, “What is the evidence?” rather than “What do I feel?” It’s not about becoming cold or detached—it’s about becoming less captured by emotionally driven distortions.

A similar idea shows up in Acceptance and Commitment Therapy (ACT), though it arrives there by a different route. Rather than challenging thoughts directly, ACT encourages a kind of psychological distance from them—what’s often called “defusion.” Thoughts and feelings are seen as transient events in the mind and body, not as commands or facts that must be obeyed. You can feel fear without being in danger. You can feel responsible without actually being responsible. And crucially, you can care about others without becoming overwhelmed by their suffering. From this perspective, the goal is not to eliminate empathy, but to hold it more lightly—to experience it without being controlled by it. In both CBT and ACT, the deeper lesson is the same: our feelings matter, but they are not sovereign. And learning to relate to them differently is often what allows us to act with greater clarity, balance, and ultimately, genuine compassion.