Autistic people do experience empathy, and the research increasingly shows that the old stereotype of autism as an “empathy deficit” fundamentally misrepresents what is happening. The reality is layered: autistic individuals often score lower on tests of cognitive empathy (reading what someone else is thinking or feeling), but their emotional responses to other people’s pain and joy are frequently intact and sometimes stronger than average. Much of what looks like missing empathy turns out to be driven by co-occurring traits, differences in emotional expression, and measurement tools that confuse recognizing emotions with actually feeling them.
What the Research Actually Finds
When researchers measure empathy in autistic people, the results depend heavily on what kind of empathy they are measuring. Cognitive empathy involves figuring out what someone else is feeling. Affective empathy involves sharing that feeling once you recognize it. These are different mental processes, and they do not always travel together.
A study using the Perth Empathy Scale found a large difference in cognitive empathy between autistic and non-autistic adults, with autistic participants reporting lower scores. For affective empathy, the picture was more mixed: autistic individuals showed a similarly large reduction in shared feelings for positive emotions, but only a moderate reduction for negative emotions. There was also much greater variability within the autistic group, meaning some autistic people scored well above the non-autistic average while others scored well below it.1PubMed Central. Empathy and Autism: Establishing the Structure and Different Manifestations of Empathy in Autistic Individuals Using the Perth Empathy Scale That spread matters. It suggests autism does not come with a single empathy profile; the range of experiences is wide.
Another study of young adults found that the autistic group had difficulty accurately reading anger in others but did not differ from non-autistic participants in broader measures of cognitive or affective empathy on the same task.2PubMed Central. Empathic Accuracy and Cognitive and Affective Empathy in Young Adults With and Without Autism Spectrum Disorder That is a specific difficulty with one emotion, not a blanket inability to empathize. Research looking at children and adults together found that while autism diagnosis predicted lower cognitive empathy performance, age was an even bigger predictor, with children across both groups recognizing emotions less accurately than adults. For emotional empathy specifically, the effect of having an autism diagnosis was small and not practically significant as a standalone factor.3Scientific Reports. Multifaceted empathy differences in children and adults with autism
The Alexithymia Factor
One of the most important findings in this area is that much of the empathy difference attributed to autism may actually be driven by alexithymia, a separate condition involving difficulty identifying and describing your own emotions. Alexithymia occurs in roughly half of autistic people, compared to about one in ten in the general population. That overlap has muddied decades of research.
A brain imaging study measured empathy-related activity in the left anterior insula, a region strongly associated with feeling another person’s pain. The researchers found that the strength of this brain signal was predicted by the person’s level of alexithymia in both autistic and non-autistic groups. After statistically controlling for alexithymia, there was no difference in empathic brain activity between the autistic and control groups.4PubMed Central. Empathic brain responses in insula are modulated by levels of alexithymia but not autism That finding is striking: it suggests the empathy differences researchers kept observing in autistic samples were being caused by the alexithymia that co-travels with autism, not by autism itself.
Behavioral data backs this up. Autistic people who also had alexithymia showed lower cognitive and affective empathy than autistic people without it.5PubMed. The Feeling of Me Feeling for You: Interoception, Alexithymia and Empathy in Autism In other words, if you removed the alexithymia from the equation, much of the empathy gap between autistic and non-autistic people would shrink or disappear. This has real-world implications: therapies targeting alexithymia, like improving emotional vocabulary and body-awareness skills, could be more helpful than interventions that treat empathy itself as the problem.
Feeling Too Much, Not Too Little
A growing body of evidence suggests that many autistic people experience something closer to empathic overload than empathic absence. When shown images of others in pain, autistic children displayed larger pupil dilation, a physiological marker of emotional arousal, compared to non-autistic children. At the same time, they showed reduced outward empathic concern and tended to avoid looking at the painful images.6PubMed. Empathy in autistic children: Emotional overarousal in response to others’ physical pain Their bodies were registering more distress, not less. The behavioral withdrawal that observers might interpret as indifference was more like turning away from a sound that is too loud.
Brain imaging in adults tells a similar story. Adults with autism showed enhanced skin conductance responses and increased neural activity in the anterior insular cortex when witnessing someone else’s pain, even though their ability to behaviorally discriminate between painful and non-painful scenarios was reduced.7PubMed Central. Autonomic and brain responses associated with empathy deficits in autism spectrum disorder The problem was not that the emotional signal was absent; it was that the brain’s system for regulating that signal appeared to function differently, with reduced ability to attenuate the response in the anterior insula.
Research on arousal regulation found that individuals with higher autistic traits showed heightened sympathetic nervous system activity when witnessing pain but had more difficulty downregulating that arousal afterward.8Heartfelt Emotion: Heightened Sympathetic Arousal and Reduced Changes in Parasympathetic Arousal during Empathy for Pain in Autism. Heartfelt Emotion: Heightened Sympathetic Arousal and Reduced Changes in Parasympathetic Arousal during Empathy for Pain in Autism This maps onto what many autistic adults describe in their own words: a high proportion of autistic people report their empathic responses as overwhelming or even distressing.9PubMed Central. Autistic People’s Experience of Empathy and the Autistic Empathy Deficit Narrative The “intense world” theory of autism proposes that social withdrawal can be triggered by overly strong sensory and emotional responses, not by the absence of those responses. One study of how people process facial expressions found its results supported this theory more strongly than the older “mind-blindness” model.10PLoS ONE. Processing of expressions by individuals with autistic traits: Empathy deficit or sensory hyper-reactivity?
The Gap Between Feeling and Showing
A persistent confusion in both research and everyday life is the assumption that if you do not look empathetic, you are not feeling empathy. Autistic people tend to express emotions differently: with less conventional facial movement, different vocal tone, or atypical body language. Non-autistic observers consistently misread these differences.
A recent study had autistic and non-autistic participants watch emotionally evocative scenarios and then recorded both how much emotion they reported feeling and how their expressions were perceived by non-autistic observers. The two groups reported comparable levels of felt emotion. But non-autistic raters perceived autistic people’s expressions as less emotional or harder to read.11PubMed Central. Discrepancies between feeling and expressing: Perceptions of autistic and non-autistic emotional expressions by non-autistic observers The gap was in interpretation, not in experience. Autistic people felt just as deeply but showed it in ways that non-autistic people did not recognize as empathic.
This expression gap may have been responsible for mischaracterizing autism as an empathy disorder in the first place. If the researchers studying empathy were themselves non-autistic and relied on behavioral displays as evidence of internal states, they would systematically undercount empathy in people whose displays did not match their expectations.
The Double Empathy Problem
The traditional framing puts the empathy difficulty entirely on the autistic person: they cannot understand what non-autistic people feel. But the “double empathy problem,” a concept developed by autistic researcher Damian Milton, argues the difficulty runs both ways. Non-autistic people are equally poor at reading autistic people’s emotions and intentions.
Experimental work supports this. A study had non-autistic adults observe recorded conversations between autistic and non-autistic people. The observers both detected and demonstrated the double empathy problem: they could see that something was off in the cross-neurotype interactions, and their own ratings showed the same bias, with reduced empathic accuracy toward the autistic participants.12PubMed Central. Non-autistic observers both detect and demonstrate the double empathy problem when evaluating interactions between autistic and non-autistic adults When autistic people interact with other autistic people, communication and rapport often flow more naturally, just as non-autistic people communicate more smoothly with each other. The breakdown is not about one side lacking empathy; it is about a mismatch in communication style.
Why Older Theories Got It Wrong
The idea that autistic people lack empathy traces partly to the “broken mirror neuron” hypothesis, which proposed that the neural system responsible for automatically imitating and understanding others’ actions was dysfunctional in autism. Research has challenged this. A study of automatic facial mimicry found that autistic individuals showed intact involuntary mimicry of emotional expressions, performing comparably to controls on congruency tasks. Their automatic, unconscious mirroring of other people’s faces worked fine.13The British Psychological Society. No, autistic people do not have a “broken” mirror neuron system – new evidence One interesting twist: in the control group, higher self-rated empathy correlated with faster mimicry, but in the autistic group, that correlation was absent. So the mimicry system was intact, but it was less connected to the person’s conscious experience of empathy.
Brain imaging comparing autistic adolescents to those with conduct disorder and callous-unemotional traits, a condition that genuinely does involve reduced emotional responsiveness, found different neural patterns between the two groups. While both showed some reduced activation in areas like the amygdala compared to typically developing controls, the specific regions affected were distinct. The autistic group showed reduced responses in the hippocampus and ventromedial prefrontal cortex, while the callous-unemotional group showed reduced responses in areas more directly tied to emotional resonance like the inferior frontal gyrus and anterior insula.14Journal of Child Psychology and Psychiatry. Different brain responses during empathy in autism spectrum disorders versus conduct disorder and callous-unemotional traits In plain terms, the brain signature of autism during empathy tasks looks fundamentally different from the brain signature of someone who genuinely does not care about others’ distress. Lumping them together under “empathy deficit” was never accurate.
Gender, Camouflaging, and Who Gets Believed
Gender adds another layer. Autistic girls and women tend to camouflage their social differences more than autistic boys and men, consciously performing social behaviors they have learned by observation rather than intuition. Research has found that camouflaging is more common in autistic females and that it correlates negatively with emotional expressivity in women but not in men.15PubMed Central. Brief Report: Sex/Gender Differences in Symptomology and Camouflaging in Adults with Autism Spectrum Disorder In practice, this means autistic women may appear more empathetic to outside observers because they are working harder to display the expected social signals, even if the internal experience is similar to autistic men who are not masking as heavily.
A study of autistic and non-autistic adolescent girls and boys found that empathic behavior was remarkably similar across groups. Autistic and non-autistic adolescents comforted an experimenter who appeared distressed at similar rates. The differences were subtle: autistic teens showed less visible emotional arousal and used different comforting styles. Autistic boys addressed the problem more often than any other group, while girls, both autistic and non-autistic, more often addressed the emotion of the person in need.16PubMed Central. Quantity and quality of empathic responding by autistic and non-autistic adolescent girls and boys The researchers interpreted this as showing that autistic adolescents do empathize and act on it, but their style of responding may prioritize practical problem-solving over emotional mirroring.
Moral Reasoning and Empathy Toward Unusual Targets
If autistic people truly lacked empathy, you might expect their moral reasoning to reflect that. It does not. A study of children with high-functioning autism found they made correct moral judgments at rates comparable to typically developing children and even applied stricter criteria for what counted as morally wrong behavior.17PubMed Central. The relationship between moral judgment and cooperation in children with high-functioning autism Where the groups differed was in how they adjusted their behavior based on a partner’s morality: typically developing children cooperated more with a “nice” partner and less with a “naughty” one, while autistic children cooperated at a steady rate regardless. That reads less like an empathy failure and more like a principled consistency, or difficulty integrating social reputation information into real-time decisions.
Autistic people also sometimes direct empathy toward targets that non-autistic people do not. Object personification, the tendency to assign feelings and personality to inanimate objects, occurs commonly among autistic individuals and may happen more often and persist later in life than in the general population.18PubMed. Object personification in autism: This paper will be very sad if you don’t read it This is not a sign of social confusion so much as an expanded circle of empathic targets. If you feel genuine distress at the thought of a stuffed animal being abandoned, your empathy system is clearly working. It is just aimed at something most people would not extend it toward.
How Culture Shapes What Gets Measured
Most empathy research on autistic people has been conducted in Western, English-speaking countries, using instruments designed by and normed on those populations. A cross-cultural comparison of autistic traits across Iran, Malaysia, Morocco, and the Netherlands found that cultural background affected how people interpreted and reported those traits.19SpringerLink / PubMed Central. Cross-Cultural Differences in the Interpretation of Autistic Traits: A Comparison Between Iran, Malaysia, Morocco, and The Netherlands What counts as an empathic response, how much emotional display is expected in a given situation, and what silence or withdrawal communicates all vary by culture. An autistic person’s empathy profile might look very different depending on where the measuring stick was made.
Even within a single culture, the tools used to assess empathy have limitations. Common measures like the Reading the Mind in the Eyes Test ask participants to identify emotional states from still photographs of people’s eyes. That format strips away context, body language, vocal tone, and situational information, all of which autistic people may rely on more heavily than non-autistic people do. The test also uses only one set of faces, limiting how generalizable its findings are. Low performance on such a task tells you something about recognizing isolated eye expressions; it tells you much less about whether someone can empathize in the complex, multi-channel world of real social interaction.
When Empathy Research Misses the Point
Brain imaging studies of empathy in autism sometimes find reduced activation in regions associated with social understanding, like the medial prefrontal cortex and the temporoparietal junction, when autistic participants view scenes of people being harmed by others.20Social Cognitive and Affective Neuroscience. Empathic arousal and social understanding in individuals with autism: evidence from fMRI and ERP measurements These regions are more closely associated with theory of mind, the ability to think about why someone did something and what they might be thinking, than with the raw emotional response to witnessing suffering. What this means is that autistic people may process the social context of a harmful scenario differently without being emotionally indifferent to the harm itself. The distinction is between understanding the social narrative and caring about the person who got hurt. They can coexist independently.
This is why the blanket claim that “autistic people lack empathy” does so much damage. It conflates the cognitive challenge of reading social cues with the emotional capacity to care. A person who struggles to identify that you are sad from your facial expression alone may still be deeply moved once they understand you are hurting. A person who does not respond to distress in the expected way, with the expected facial expression, at the expected moment, may still be experiencing empathic pain so intense they have to look away. The research, taken as a whole, points not to absent empathy in autism but to empathy that takes a different route and sometimes arrives in a form that non-autistic people do not recognize.