Autistic Brains Process Faces Differently, And Now We Know Why

Ever wonder what’s really going on when someone looks at a face? For neurotypical brains, it's a finely tuned symphony of recognition, expression, and identity. For autistic brains, it turns out, it’s a very different concert. New research suggests these distinct brain signals aren't just different, they also don't specialize with age in the same way, offering a clearer picture of why interpreting faces can be such a puzzle for some autistic individuals.

Published in Nature Mental Health, this large-scale study gives us a satellite view instead of just a street-level glimpse, and the implications are significant.

The Brain's Facial Recognition Software

When a neurotypical child sees a face, their brain quickly gets to work: Whose face? What emotion? Is that an eyebrow raise of surprise or suspicion? Autistic people, however, often avoid eye contact, and might struggle with remembering faces or recognizing emotions. Past research from the early 2000s hinted that brain areas for face processing were less active in autistic individuals, but the scope was limited.

James McPartland, a professor at the Child Study Center at Yale, has been using electroencephalography (EEG) — essentially, recording the brain’s electrical chatter — to dig deeper. But here’s the kicker: earlier EEG studies were like watching city traffic from one intersection. McPartland’s team, with data from about 400 autistic children, took a different approach.

Instead of focusing on a few hotspots, they analyzed signals from all 128 electrodes on the scalp. Then, they unleashed machine learning to see if brain activity patterns could predict what each child was looking at. The result? The system could more easily tell when a neurotypical child was looking at a face compared to an autistic child. As Jason Griffin, the first author, put it, the neural activity in autistic children was simply “not as distinct.”

And here’s where age comes in: for neurotypical children, this ability to predict what they were viewing improved with age, the brain signals becoming more refined. For autistic children, that age-related refinement was conspicuously absent. Their face-specific processing, it seems, just doesn't follow the same developmental script.

Earlier Than We Thought

This study also shifts our understanding of when these differences in face processing begin. Previous research often fixated on the N170 biomarker, a burst of brain activity about 170 milliseconds after seeing a face. It's often delayed in autistic people. But by looking at face processing more broadly, McPartland’s team found that differences in autism appear even earlier than the N170.

This isn't about pathologizing autism. Many neurodivergent people rightly see autism as a different way of experiencing the world, not something to be “fixed.” However, for some, the difficulty in interpreting faces can make daily life a genuine challenge. Think about it: a playground, a job interview, navigating social cues — all suddenly become more complex when facial information is harder to decipher.

This new research suggests these face-processing differences are intrinsic to autism, not just a consequence of later experiences. And that means early support for interpreting facial emotions might be most helpful when these differences first emerge in childhood.

Currently, autism diagnosis relies on behavior because there are no biological assessment tools. But a clearer understanding of these face-processing differences could pave the way for new diagnostic tools, or at least help identify who might benefit most from interventions like social skills therapy. McPartland sees this as a crucial first step in introducing a new biomarker. Which, if you think about it, is both impressive and slightly terrifying in its potential.