A new exploratory study published in the Journal of Autism and Developmental Disorders finds that autistic adolescents and adults make broadly similar moral decisions to their neurotypical peers, but arrive at those judgments with subtly different patterns of contextual reasoning. The research, led by Margherita Attanasio and Monica Mazza and colleagues at e-Campus University and the University of L’Aquila in Italy, examined how autistic and non-autistic individuals weigh harm, intentionality and personal risk when confronting classic moral dilemmas—and probed whether the so-called “dark” personality trait of Machiavellianism plays any role in the differences.
Moral decision-making has long fascinated psychologists because it pits two competing systems against each other: a cold, calculating tendency to maximize outcomes (the utilitarian response) and a hot, emotionally driven resistance to directly causing harm to another person (the deontological response). In the canonical trolley problem, most people say it is permissible to divert a runaway trolley to kill one person instead of five—a utilitarian choice made at a distance, where harm is an incidental side effect. Yet most of those same people refuse the equivalent choice in the footbridge variant, where one person must be physically pushed to their death—harm used as an instrument to achieve the goal. Neuroimaging work by Joshua Greene and colleagues, published in Science in 2001, showed that these “personal,” high-conflict dilemmas engage emotional and social brain regions more strongly, suggesting that emotional aversion to up-close harm drives the refusal to sacrifice one to save many.
Whether autistic people show this same profile has been debated for decades. Early theoretical accounts proposed that autism involves impairments in empathy and Theory of Mind (ToM)—the capacity to attribute mental states such as beliefs and intentions to others—which might blunt the emotional aversion to harm and produce more utilitarian, even coldly calculating judgments. Some studies reported exactly that, fueling a persistent and harmful stereotype that autistic people lack moral feeling. More recent systematic reviews, including work by Dempsey and colleagues, have suggested that basic moral judgment is largely intact in autism, though findings remain mixed and the literature suffers from small samples and inconsistent methods. The new Italian study adds an unusually detailed data point to this debate by manipulating two specific features of moral dilemmas: whether the harm is incidental or instrumental, and whether the decision-maker is personally at risk.
The researchers recruited twenty autistic adolescents and adults (mean age 21.60, standard deviation 8.10, with a mean of 12.40 years of education) and twenty-nine neurotypical participants (mean age 22.28, standard deviation 2.8, with 14.38 years of education). Autistic participants were assessed using the Autism Diagnostic Observation Schedule, Second Edition (ADOS-2). All participants read a battery of moral dilemmas drawn from validated Italian normed sets and rated the moral acceptability of utilitarian courses of action. The dilemma set systematically varied along two dimensions: incidental dilemmas, in which harm befalls someone as a side effect of an action that produces a greater good, versus instrumental dilemmas, in which harm to the victim is the means by which the good outcome is achieved; and dilemmas in which the protagonist’s own life was at stake versus those in which only others’ lives were threatened.
To examine individual differences that might underpin moral choices, participants also completed measures of empathy (the Empathy Quotient), advanced Theory of Mind, and Machiavellianism—a personality construct characterized by cynicism, strategic manipulation and emotional detachment, classically measured with instruments derived from the Christie and Geis Mach-IV scale. The researchers then analyzed the data using mixed-design repeated-measures ANOVAs, a statistical approach that allows within-subject factors (dilemma type and personal risk) to be evaluated alongside the between-subjects factor of group membership. Exploratory correlation analyses were conducted to see how ToM, empathy and Machiavellianism related to utilitarian responding within each group.
The headline finding was one of similarity. Both groups showed the classic signature of moral cognition: significantly higher acceptability ratings for utilitarian actions in incidental dilemmas than in instrumental ones, and a drop in utilitarian choices when harm became the instrument of salvation. In other words, autistic participants, like neurotypical ones, found it more acceptable to let harm happen as a side effect than to use a person as a means to an end—a pattern consistent with the deontological constraints documented across the general population. The authors emphasize that overall moral outcomes in the autistic group were “broadly comparable” to those of the neurotypical group, a result that directly counters lingering assumptions of moral deficit in autism.
But the fine structure of the data revealed group differences. The reduction in utilitarian responding from incidental to instrumental dilemmas, while present in both groups, was less pronounced among autistic participants. More strikingly, personal risk involvement shifted judgments in opposite ways: neurotypical participants became more utilitarian when their own lives were on the line, whereas personal risk did not significantly influence the choices of autistic participants. This suggests that the self-preservation calculus that tilts neurotypical moral judgment toward sacrifice of others may operate differently—or be integrated differently—when the decision-maker is autistic.
On the social cognition measures, the autistic group reported lower empathy and lower Theory of Mind abilities than the neurotypical group, replicating well-established group differences. Yet Machiavellianism scores did not differ between groups—a result with real significance for the stereotype question. If autistic moral reasoning were driven by cold, manipulative indifference, one would expect elevated Machiavellianism; instead, the groups were indistinguishable on this trait, indicating that any differences in utilitarian responding cannot be attributed to darker personality dispositions.
The exploratory correlations added a further layer of nuance. Rather than showing uniform relationships across the board, the analyses suggested that ToM and Machiavellianism related to utilitarian choices only in specific dilemma conditions, and in group-specific ways: relationships between these social cognitive measures and utilitarian responding emerged within each group, but tied to different dilemma contexts. The authors interpret this as evidence that autistic and neurotypical individuals may integrate information about harm intentionality, personal involvement and normative constraints through partly different cognitive routes, even when the endpoints of their decisions look alike.
The study’s methodological approach deserves attention. By crossing the incidental/instrumental distinction with personal risk, the design separates two factors that are usually confounded in the trolley-and-footbridge paradigm. The incidental versus instrumental manipulation isolates the role of intentionality—whether harm is foreseen or intended—while the personal risk manipulation isolates self-interest, a factor shown in prior work to increase utilitarian responding in the general population. The finding that these two factors behave differently in autism provides a more precise picture than a simple “autistic people are more utilitarian” headline could offer. It also aligns with a growing literature suggesting that alexithymia—difficulty identifying and describing one’s own emotions—rather than autism itself may explain atypical moral judgments in some autistic individuals, as Patil and colleagues reported in Scientific Reports in 2016.
The authors are careful to frame their conclusions within the limits of an exploratory design. The sample was modest in size, with unequal group numbers, and the correlation analyses were explicitly exploratory rather than confirmatory, meaning they generate hypotheses rather than establish robust effects. Education levels also differed slightly between groups, and the autistic group showed greater age variability. These constraints temper strong generalizations. Nevertheless, the convergent picture—intact core moral judgment, intact aversion to instrumental harm, reduced sensitivity to personal risk, and no elevation in Machiavellianism—adds to a mounting body of evidence that the moral minds of autistic people are not deficient, merely differently calibrated.
Why might personal risk fail to boost utilitarian responding in autism? The authors suggest that context-dependent patterns may reflect differences in how personal involvement and normative constraints are integrated into the decision process. One possibility is that neurotypical self-interest operates through emotional and interoceptive signals—visceral fear, anticipated regret—which may register differently in autistic cognition, a question increasingly studied through the lens of interoception and alexithymia in autism research. Another possibility relates to rule-based reasoning: if autistic participants anchor judgments more firmly in explicit moral rules than in self-serving emotional pulls, their choices would remain stable regardless of whether their own survival is at stake. The study cannot adjudicate between these accounts, but it sets up clear hypotheses for future work with larger samples and, potentially, virtual reality paradigms that make dilemma contexts more immersive, as recent reviews of moral dilemma methodology have proposed.
The findings also carry practical implications. Clinicians and educators sometimes encounter assumptions that autistic young people struggle with moral reasoning, which can affect how social conflicts involving autistic individuals are interpreted. This study suggests that when an autistic person endorses a utilitarian solution to a dilemma, or declines to shift their judgment under self-preservative pressure, this reflects a different integration of contextual information rather than a lack of moral feeling or an excess of cunning. Given that Machiavellianism did not differ between groups, the research offers no support for the harmful trope of the “calculating” autistic mind.
In the broader landscape of moral psychology, the study contributes to a reframing that has been gathering momentum: moral judgment differences across neurotypes are best understood not as deficits but as variations in how intention, causality, emotion and self-interest are weighted. As the authors conclude, moral decision-making in autistic individuals was broadly comparable to that of neurotypical individuals in overall outcomes, while the context-dependent patterns they observed point to meaningful differences in the underlying architecture of moral cognition. For a field still working to disentangle autism from its stereotypes, that is a finding worth pushing a trolley for.
Subject of Research: Moral decision-making, Theory of Mind, empathy and Machiavellianism in autistic versus neurotypical adolescents and adults, examined through moral dilemmas varying in harm intentionality and personal risk
Subject of Research: Medicine
Article Title: Moral Decision-Making, Machiavellianism, and Social Cognition in Autism: An Exploratory Study
Article References: Attanasio, M., Mazza, M., Le Donne, I., Covone, N., & Valenti, M. (2026). Moral Decision-Making, Machiavellianism, and Social Cognition in Autism: An Exploratory Study. Journal of Autism and Developmental Disorders. https://doi.org/10.1007/s10803-026-07509-8
Image Credits: AI Generated
DOI: 10.1007/s10803-026-07509-8
Keywords: Autism spectrum disorder, Moral dilemma, Moral decision-making, Machiavellianism, Theory of mind, Empathy, Utilitarian judgment, Personal risk
Cite Scienmag News
APA MLA Chicago
Glenn Wilkins. (September 4, 2026). Study Explores Moral Choices, Machiavellianism, and Social Cognition in Autism. Scienmag. https://scienmag.com/study-explores-moral-choices-machiavellianism-and-social-cognition-in-autism/
Glenn Wilkins. “Study Explores Moral Choices, Machiavellianism, and Social Cognition in Autism.” Scienmag, 4 September 2026, https://scienmag.com/study-explores-moral-choices-machiavellianism-and-social-cognition-in-autism/. Accessed 4 September 2026.
Glenn Wilkins. “Study Explores Moral Choices, Machiavellianism, and Social Cognition in Autism.” Scienmag. September 4, 2026. https://scienmag.com/study-explores-moral-choices-machiavellianism-and-social-cognition-in-autism/
Copy citation Download RIS
Tags: autismAutism and moral decision-makingautism research in moral psychologyautistic adolescents and adultscontextual reasoning differences in autismcontextual reasoning in autismharm and intentionality assessmentharm and intentionality in moral choicesimpact of personality traits on moral reasoningMachiavellianism and autismMachiavellianism and personality traitsmoral decision-making in autismmoral dilemmas and autismmoral dilemmas and ethical reasoningmoral reasoning patternspsychological factors influencing moral decisionspsychological traits in autismsocial and emotional processing in moral judgmentssocial cognition and personality traitssocial cognition in autismsocial cognition in autistic individualsutilitarian vs deontological reasoningutilitarian vs deontological responses
