Simone Grassini

The increasing presence and quality of AI-generated human faces requires a clearer understanding of how humans perceive these faces. Recent findings indicate AI-generated human faces are rarely effectively distinguished from biological human faces. In this EEG study, 28 participants viewed images of humans, AI-generated humans, robots, and dolls displaying happy, fearful, and neutral expressions. Event-related potentials associated with face processing (N170), expectancy violation and uncanniness (N400), emotionally driven attention (EPN) and sustained attention (LPP) were examined across conditions. Additionally, a questionnaire regarding perceptions of the faces was administered after EEG recordings. This poster presents analyses, some of which explorative, related only to real human and AI-generated human faces. Results yielded significant differences between real and AI-generated human faces in three early ERPs N170, N400, and EPN, but not for the later LPP. Significant differences emerged for expressiveness in emotionally expressive and neutral human and AI-generated faces. Sense of self and inner consciousness was also significantly different across faces. Perceived intelligence and appeal of faces were non-significant. Together, these mixed results indicate that there may be other cognitive processes involved in the processing of human and AI-generated faces than the knowledge or suspicion of realness. Results also indicate that faces are at early processing stages differentiated.

Growing concerns about the psychological impact of generative AI have highlighted the need for better measurement tools. Existing scales are limited by narrow samples and a failure to distinguish anxiety from stress. This research introduces the Artificial Intelligence Stress and Anxiety Scale (AISAS), a psychometrically validated instrument assessing both constructs in the general population. Development followed best-practice guidelines across two preregistered studies using representative UK adult samples for age, gender, and ethnicity. An initial 62-item pool was refined via expert ratings and a Content Validity Index. Exploratory factor analysis (Study 1, N = 301) identified four factors: job-related concerns, AI adaptation stress, privacy concerns, and existential/consciousness concerns, explaining 72% of variance. The adaptation stress dimension was retained as a brief two-item indicator ( r = .68, Spearman-Brown ρ = 0.81) rather than a fully developed subscale. The general scale showed excellent internal consistency (α = 0.92), full gender measurement invariance, and preliminary evidence of convergent, discriminant, and criterion-related validity. Specific dimensions differentially predicted AI usage frequency and behavioral intention. Study 2 ( N = 324) confirmed the factor structure’s stability across an independent sample. Descriptive findings suggest privacy concerns are common, while stress related to adapting to and learning about AI is relatively infrequent. Overall, the AISAS provides a reliable, multidimensional tool for researchers and practitioners assessing psychological responses to artificial intelligence.

Extending conventional health and well-being services with technology has proven to be acceptable and effective. However, implementing such technology in routine care requires contextual tailoring to ensure successful uptake. The Unified Theory of Acceptance and Use of Technology (UTAUT2) is an established model with predictors and moderators of intention to use technology and use behavior. The current study aimed to design a concise screener for technology acceptance in the health and well-being field based on UTAUT2 predictors and to provide a preliminary evaluation of the instrument. The 11-item UTAUT2-Screener includes 7 predictors, i.e., performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation, price value, and habit, as well as three moderators and behavioral intention. A within-subject repeated measures design was used. Spearman’s correlations were used to test the convergent validity concurrent validity, and the test-retest reliability. Analyses on 184 individuals from the general population demonstrated a moderate association with the original UTAUT2 (mean ρ = 0.63), with some variability across the subscales. More specifically, facilitating conditions had the weakest, but still moderate association, while hedonic motivation exhibited the strongest correlation. As an instrument with single-item scales, the UTAUT2-Screener is provided for use in contexts where only brief measures are feasible and researchers or clinicians are willing to accept the reduced psychometric robustness that comes with shorter assessments. The screener provides insights into specific barriers for technology adoption and can consequently support the development of tailored acceptance-promoting interventions, hereby supporting the uptake of digital health and well-being services in practice.

Sleep plays an important role in affective processing and affect regulation during wakefulness. Dreaming—the subjective experiences occurring during sleep—has been suggested to reflect, or even contribute to, waking affect regulation. However, the role of dream affect in waking affect regulation has rarely been studied directly. Here, we investigated how dream affect relates to waking affect reactivity and regulation at the between-person level. Over 7 days, 87 participants (18–70 years, M = 27.23, SD = 11.75; 76 females) kept dream journals and rated their dream affect and sleep quality each morning upon awakening. They then completed an affect reactivity and regulation task in the laboratory while their brain activity was measured using electroencephalography. Participants were shown negative and neutral pictures and instructed to either (1) look at the pictures and experience affect elicited by the pictures or (2) downregulate their affective responses. We analyzed subjective ratings of valence and arousal as well as the late positive potential (LPP), a well-validated index of affect reactivity and regulation. Multiple linear regression analyses showed that participants with higher average levels of dream fear and good sleep quality displayed stronger affect reactivity (indexed by larger LPP amplitudes) to negative pictures. No significant relationship emerged between dream affect and subjective ratings of affect reactivity or between dream affect and waking affect regulation. These results support, at the between-person level, continuity between dream affect and objective, neurophysiological affect reactivity in wakefulness but not continuity in subjective affective experiences.

Love is a near-universal human experience. However, while it transcends cultural boundaries, its intensity varies significantly across individuals and societies. We hypothesize that this variability is partly driven by socio-cultural factors, particularly individualism—a cultural orientation prioritizing personal goals over group interests. As individualism continues to rise globally, its implications for romantic relationships remain underexplored. Here, we address this by examining the link between individualism and love in two large-scale cross-cultural studies. In Study 1, we collected data from a convenience sample of 61,067 partnered individuals from 81 countries. In Study 2, we selected nationally representative samples from 50 countries, with controlled distributions of gender, age, and residential area (N = 6,620). The results were consistent across both studies—as individualism increased, both men and women reported lower levels of experienced love. In Study 2, individualism was additionally linked to a lower expressed desire for love, suggesting that individualism may erode the perceived importance of love. As people become increasingly individualistic, our findings provoke questions about the future of love. They also carry implications for societal challenges such as rising divorce rates, fertility issues, and depression—all of which have been linked to unfulfilled or unhappy love.