The costs of being seen: disclosure and safety

Non-autistic observers tend to form less favorable first impressions of autistic people than of non-autistic people. A 2025 meta-analysis of 21 articles found the gap in every format tested except written transcripts, usually moderate to large. It was larger for how attractive people seemed as company and for how they came across socially than for judgments of personality.¹ There was some evidence that observers who knew more about autism, or had better past contact with autistic people, formed better impressions.¹

In three studies from one group, these judgments formed within seconds, did not change with longer viewing, held whether the people rated were children or adults, and went with less wish to interact. They disappeared when raters read only what had been said.² Another group's study of mock job interviews found the same split: non-autistic candidates were rated higher on video, and autistic candidates higher on transcripts.³ A Canadian study of 346 students rating autistic children found the gap too, mainly in clips with sound.⁴ In these studies the gap appeared only when raters could see or hear the people they rated.

When 505 non-autistic raters judged 20 autistic adults, differences between raters' judgments, especially of whether they wanted to interact, came more from the raters' stigma and knowledge than from the autistic adults; ratings of awkwardness were the exception.⁵ In a study with 32 autistic raters, they too rated autistic adults less favorably than non-autistic adults, but this did not reduce their interest in meeting them.⁶ Two studies found autistic women rated more favorably than autistic men.^9,7 In one, using 10-second video clips of 10 autistic women and 10 autistic men, the women were still rated less favorably than non-autistic women and men; on transcripts, autistic men were rated more favorably for what they said.⁷

In experiments, telling non-autistic observers that someone is autistic usually improves how they rate and respond to that person, though in hiring studies it did not on its own make autistic candidates as likely to be chosen.^{9,10,11,12,14} After diagnosis describes the first of these studies; others extend them. In three Israeli studies with one autistic storyteller, disclosure raised empathy in two studies, with a similar trend in the third, and raised willingness to work together in the two that asked; effects on wider social interest were inconsistent.¹⁰ In 170 Irish participants reading about a co-worker with autistic traits, a label reduced the wish to keep social distance.¹¹

Part of the effect belongs to the label itself. In an Australian study using one autistic and one non-autistic man, raters scored both candidates higher when both were labeled autistic, and still rated and "hired" the non-autistic candidate more.¹² In a vignette study, 120 students reported more positive and less negative feelings when the behavior was attributed to a clinical label (autism, Asperger's or schizophrenia) than to a typical student, and the three labels did not differ.¹³ Disclosure made impressions worse among raters high in autism stigma, and did not change autistic raters' impressions.^5,6

One research group, the same that ran the transcript study above, found that neither disclosure nor neurodiversity training alone closed the gap in hiring. Together, they made autistic candidates as likely to be hired as non-autistic ones, though they were still rated lower on some social traits, and the effect held when training came two weeks or two months before.^14,15

Real-world outcomes of disclosure at work are mixed: a systematic review of 26 studies found wide variation, with acceptance and adjustments for some and stigma and discrimination for others.¹⁶ After diagnosis covers those studies.

Among 131 employed autistic adults in the US, surveyed at the start of a trial of a workplace training program, 86% camouflaged often or sometimes at work. Camouflaging was higher in women, non-binary and transgender people than in men. In this sample no association was found with race, ethnicity, sexual orientation or whether people had disclosed.¹⁷ In written answers, those who had not disclosed described active coping strategies at more than twice the rate of those who had.¹⁷

Among 123 autistic students in the US, willingness to tell a professor was predicted by how safe disclosure felt at their university and by how much they needed adjustments; most wished they could disclose without consequences. White, cisgender and heterosexual students were more likely to disclose than students of color and LGBTQIA+ students, though White students were overrepresented in the sample.¹⁸

At UK universities, support frequently depended on diagnostic evidence and on staff discretion. A Freedom of Information study of 142 publicly funded universities, answered by 125 of them, found that half kept records on formally diagnosed autistic students, 16% on students seeking a diagnosis and 7% on students who identified as autistic without wanting a diagnosis; only 2% had a policy for students without one. Those percentages describe which students universities keep records on; the study's wider finding is that staff judgment and discretion mostly decided whether support and adjustments were given, and that support often turned on diagnostic evidence even though disability law ties it to need.⁴² The study reports what universities said about their own practice, so it describes institutional policy rather than students' experience of it.

A global online survey of 576 autistic people, mostly highly educated, found preliminary evidence that White respondents in Western countries were more likely to be employed. Among those in work, women and non-binary people felt less included, and feeling accommodated went with a sense of belonging.¹⁹

Among autistic people who had been pregnant, many report hiding their traits in healthcare. Of 193 in the UK, about half formally diagnosed, over 80% reported anxiety and masking during healthcare always or most of the time. In open-text answers, the main reason given for not using an autism health passport was a belief that health professionals would discriminate against autistic patients.²⁰

In an international online survey shared through autistic networks and social media, 80% of 507 autistic adults reported difficulty visiting a GP, against 37% of 157 non-autistic adults. Not feeling understood and difficulty communicating with the doctor were among the barriers most often named, and barriers went with self-reported harms such as untreated conditions and late presentation.²¹ Written accounts from the same survey program describe doubt from doctors, fear, and avoiding healthcare.²²

Some autistic people carry more than one stigmatized identity, and some studies report more masking among them. In 462 autistic adults matched for sex assigned at birth and age, sexual minority adults reported more masking than heterosexual adults.²³ Thirteen autistic transgender, non-binary and gender-diverse adults described constantly weighing whether it was safe to disclose their identities.²⁴ In a US survey, autistic cis women and gender-diverse adults were more worried than cis men that their autistic traits would be misread as dangerous by police.²⁵

In interviews with 102 autistic people, being hurt by people they knew was often treated as normal, and many described masking to stay safe, which was exhausting. Those who sought help often met professionals, police among them, who knew little about autism.²⁶

In a nationally representative US sample of autistic young people surveyed at 14 to 15 and again at 21 to 22, about 20% had been stopped and questioned by police by age 21, and nearly 5% had been arrested. That paper has no non-autistic comparison, so it shows how often contact happened, not whether it happened more than for other young people. More externalizing behavior went with more contact, and being female with less.²⁷

A UK cohort supplies the comparison the US sample lacks. In the Millennium Cohort Study, 11,859 young people were followed at ages 14 and 17, 503 of them with an autism diagnosis, and 17.6% of the autistic group reported a police warning by age 14 against 8.9% of the others, about twice the risk before other factors were taken into account. After full adjustment for the study's covariates, including delinquency, school disengagement and poverty, the difference was no longer significant; the authors caution that the sample was not powered for autistic youth as a subgroup, so this null finding should be read with care. Arrest rates did not differ in either model, and after adjustment autistic youth were less likely to have been stopped. These figures are about how common contact is, not about how visible traits change an encounter.⁴³ The same study found that contact went with worse mental health for autistic than non-autistic adolescents: among those stopped, autistic adolescents had higher odds of short-term self-harm, and among those given a warning, worse well-being and more symptoms of depression at age 17, though the estimates were imprecise.⁴³ In an editorial on the study, Yates Flanagan, Miller and Parish-Morris ask whether such associations are specific to autism and say more research is needed.⁴⁴

Experiences of that contact are often poor. In England and Wales, 42% of 394 police officers were satisfied with how they had worked with autistic people, and most of the 31 autistic adults and 49 parents surveyed were dissatisfied with their experience of the police.²⁸ Twelve autistic people held in custody described struggling to understand their arrest and to cope with the setting; a wish to get out shaped their decisions and how they handled the interview.²⁹ At six London police custody centers, screening flagged about 5% of arrestees who had no existing diagnosis as warranting further assessment for autism, on a short questionnaire rather than a diagnosis, which suggests that some difficulties go unrecognized.⁴⁵

How a person comes across can shape how readily they are judged. In two filmed-interview experiments, observers rated autistic male suspects, and autistic suspects who hesitated more in speech, more harshly. The authors trace this to the observers' expectations being violated, which went with harsher ratings whatever the suspect's diagnosis.⁸ That study rates how observers judge filmed interviews, not what happens in an encounter.

No study found as of October 2026 has measured whether showing autistic traits openly, rather than masking them, changes the outcome of a police encounter.

It has been argued, with a focus on Black autistic women, that unmasking is "a privilege not granted to those beyond the margins".³⁰ Caregiver reports fit this concern. In one interview study of 43 Black caregivers, mostly mothers, of Black autistic young people (average age 16.5, mostly boys), caregivers' concerns about police included autistic behavior being inappropriately policed and an explicit threat of harm and murder; two papers report this one group.^31,32 A critical autoethnography by one autistic multiracial Latina describes "dual masking": hiding autistic traits while also assimilating to Whiteness.³³ In 56 Latino autistic young adults enrolled in a therapy program, acculturation stress and traditional gender roles were linked to how people camouflaged.³⁴

The direct evidence is sparse and does not all point one way. Black autistic adults are largely absent from autism research.³⁵ A small online survey of 32 autistic and 30 non-autistic Black young adults gives a mixed picture: the autistic group reported less everyday discrimination, but named being autistic as well as race as a source of it, and 12 to 30% named race and disability together.³⁶ It measured discrimination in general, not reactions to visible traits. In a first-impressions experiment with 15 Black and 14 White autistic adults, non-autistic raters rated the Black autistic adults as more likeable and trustworthy. White autistic men were rated less favorably than autistic participants who were not men, while Black autistic men were rated more favorably on some items.³⁷ As its abstract describes it, only autistic adults were rated, so it cannot show whether the gap between autistic and non-autistic people differs by race. No study found as of October 2026 has measured what happens when Black autistic adults unmask.

A systematic review of 27 studies found that autistic people may manage stigma by concealing and camouflaging, by telling selectively and speaking up for themselves, and by reframing what autism means to them. Evidence on how well each strategy works was limited and mixed, and the studies covered a narrow range of intellectual ability.³⁸ One conceptual analysis argues that masking is a response to stigma and not always a choice.³⁹

The only two studies of unmasking itself describe limits from outside and from within. In interviews with 14 US college students, unmasking relieved stress and felt more authentic, but raised worries about backlash unless friends or other unmaskers backed it, and it was difficult or impossible where masking was deeply ingrained. In a survey of 214 students, some self-identified, finding unmasking hard went with more anxiety, and unmasking with family went with less anxiety and depression. The author traces its drawbacks to outside stigma.⁴⁰ In 29 Swedish autistic adults, mostly women and 17 with ADHD as well, managing the consequences of unmasking was one stage of a circular process, and supportive surroundings mattered.⁴¹

Autistic writers mostly treat telling people as a choice made person by person. Devon Price wishes every autistic person could be "loudly and visibly disabled in every area of their life", but calls that "unrealistic and oversimplified" because "we also each know our own circumstances best".^c1 Megan Anna Neff writes that disclosure "doesn't have to be an all-or-nothing approach": telling part of it lets a person gauge the response, and lets what they share "match the pace of closeness in the relationship".^c5 For work, Daniel M. Jones writes: "You only need to disclose your issues if you are in need of access."^c8 Eric Garcia, an autistic journalist, chose not to disclose at his first job, and writes that autistic people who can pass as non-autistic often find it easier not to.^c9

Daniel Burford, writing for adults diagnosed late, offers scripts of four sentences at most for a partner, a parent, a friend or a colleague: "Long explanations invite debate, and you don't want a debate, you want the fact to land and the day to continue." Where the aim is a practical change, Burford advises asking for it first and explaining only if it is refused.^c12 Luke Beardon, a practitioner who writes for autistic adults, suggests weighing whether telling will raise or lower anxiety, and telling first the people who may already suspect or are likely to accept it, which "can be such a lovely and affirming experience".^c13

Writers describe the reactions they meet. Sol Smith writes that when an adult comes out as autistic, "the initial reaction is almost always disbelief", and that people offer as counter-evidence "hard-fought parts of your masked identity" such as eye contact and charm.^c7 Price tells of a woman whose new partner said she was too "pretty and poised" to be autistic, and argues that disclosure changes someone's understanding only within "a mutually respectful, trusting relationship".^c1 Once trusted friends knew, Garcia writes, "I no longer had to worry about masking or pretending to act more neurotypical."^c10 Writers also recount bad outcomes at work: Jones describes a friend whose work "dried up" after she told employers she was autistic, and Price a friend whose boss threatened to fire her for acting "Autistically".^c8,c11

Melanie Yergeau, an autistic scholar of rhetoric, argues that "autism disclosure is often agonistic, expectant of allistic refutation": saying one is autistic is taken as an invitation for others to assess the claim, and the very ability to say it is read as evidence of not being autistic, or not really. In disclosing, Yergeau writes, autistic people are treated as "both too autistic and not autistic enough".^c16

Price calls the most open approach "radical visibility", which "is self-advocacy, as well as self-expression". Price adds that reflexes such as people-pleasing exist to protect autistic people, and "there is no shame in having them".^c2 Writers also stress that openness depends on safety. Neff writes: "To safely unmask, you must first feel safe." Neff notes that many BIPOC autistic people navigate "the double layer of masking and code-switching".^c6 Price writes that some autistic people "are in such perilous positions that opening up is too dangerous", and may keep their masks on except in small pockets of acceptance.^c3 Amy Pearson and Kieran Rose argue that "the notion of unmasking is embedded in privilege": it may be safer for a white autistic woman to change these behaviors than for a Black autistic woman, and urging people to unmask when society does not make it safe can end up blaming them for how they are treated.^c14 In a chapter on Black and brown autistic people, Price cites a figure that about half of the people killed by police are disabled; as noted above, no peer-reviewed study found for this entry measures it.^c4

Coby Khodosh, who spells on a letterboard and was nine when his chapter in Leaders Around Me appeared, writes that his letterboard "made me visible to others". He asks autistic people to be seen in public spaces, "happily spelling and living our lives": "We need to do it unapologetically, expecting acceptance, without asking for permission to exist, and without the fear of rejection."^c15