The growing integration of artificial intelligence (AI) into decision-making processes has raised important concerns about fairness and trust. This research investigates how the presence of AI, human collaboration, and outcome favorability shape perceptions of fairness and trust in decision-making in a college-admission scenario. Across two experimental studies, we examine decision-making scenarios involving AI-only, human-AI collaboration, and human-only agents, under decision outcomes that were either favorable or unfavorable. Study 1 demonstrates that human involvement, either alone or in collaboration with AI, enhances fairness and trust relative to AI-only decisions, with fairness partially explaining trust differences. Study 2 extends these findings by showing that outcome favorability strongly influences fairness and trust, often outweighing the effects of who made the decisions. Notably, while human-only decisions elicited the highest trust under favorable outcomes, trust differences across agents diminished when outcomes were unfavorable. These findings highlight the complex interplay between decision-making agents and outcome favorability, underscoring the importance of integrating human involvement and managing outcome expectations to foster fairness and trust in AI-driven decision-making systems.

This study investigates people's perceptions of AI decision-making as compared to human decision-making within the job application context.It takes into account both favorable and unfavorable outcomes, employing a 2 � 2 experimental design (decision-making agent: AI algorithm vs. human; outcome: favorable vs. unfavorable).Upon evaluating a job seeker's suitability for a position, participants viewed algorithmic decisions as fairer, more competent, more trustworthy, and more useful than those made by humans.Interestingly, when a candidate was deemed unsuitable for hiring, people reacted more negatively to the verdict given by a human than to the same judgment offered by AI.Moreover, participants credited algorithmic decisions with greater sensitivity to both quantitative and qualitative qualifications, thus indicating algorithmic appreciation.Our findings shed light on the psychological basis of perceptions surrounding Algorithmic Decision-Making (ADM) and the responses to the decisions rendered by ADM systems.

Well-being apps promise to improve people's lives. Yet evidence shows that the data-hungry app culture that contextualizes well-being apps normalizes the user experience of affective discomfort. This apparent contradiction raises a difficult question: Is it responsible to ask people to improve their well-being by engaging further with an app culture that normalizes affective discomfort? We approached this question by deploying an online, scenario-based survey (n=688) about a fictional, but realistic well-being app called "Thalia." Thalia represents a novel class of well-being apps that are envisioned to: (i) utilize AI-driven facial recognition and analysis; and (ii) collect data for use in medical contexts. We found that people perceived Thalia to be affectively discomfiting even as they judged Thalia to be beneficial. Such findings imply a trade-off between 'better living through technology' and the negative affective implications of surveillance capitalistic app culture. Such a trade-off necessitates high-level analysis of just what "well-being" means in the context of contemporary app culture. Through analysis and discussion, we explore a troubling interplay between novel well-being apps and affective discomfort that requires careful attention from HCI researchers if human-centered well-being -- flourishing -- is truly what our products are intended to foster.

As AI-enhanced technologies become common in a variety of domains, there is an increasing need to define and examine the trust that users have in such technologies. Given the progress in the development of AI, a correspondingly sophisticated understanding of trust in the technology is required. This paper addresses this need by explaining the role of trust in the intention to use AI technologies. Study 1 examined the role of trust in the use of AI voice assistants based on survey responses from college students. A path analysis confirmed that trust had a significant effect the on intention to use AI, which operated through perceived usefulness and participants’ attitude toward voice assistants. In Study 2, using data from a representative sample of the U.S. population, different dimensions of trust were examined using exploratory factor analysis, which yielded two dimensions: human-like trust and functionality trust. The results of the path analyses from Study 1 were replicated in Study 2, confirming the indirect effect of trust and the effects of perceived usefulness, ease of use, and attitude on intention to use. Further, both dimensions of trust shared a similar pattern of effects within the model, with functionality-related trust exhibiting a greater total impact on usage intention than human-like trust. Overall, the role of trust in the acceptance of AI technologies was significant across both studies. This research contributes to the advancement and application of the TAM in AI-related applications and offers a multidimensional measure of trust that can be utilized in the future study of trustworthy AI.

Automatic speaker recognition algorithms typically characterize speech audio using short-term spectral features that encode the physiological and anatomical aspects of speech production. Such algorithms do not fully capitalize on speaker-dependent characteristics present in behavioral speech features. In this work, we propose a prosody encoding network called DeepTalk for extracting vocal style features directly from raw audio data. The DeepTalk method outperforms several state-of-the-art speaker recognition systems across multiple challenging datasets. The speaker recognition performance is further improved by combining DeepTalk with a state-of-the-art physiological speech feature-based speaker recognition system. We also integrate DeepTalk into a current state-of-the-art speech synthesizer to generate synthetic speech. A detailed analysis of the synthetic speech shows that the DeepTalk captures F0 contours essential for vocal style modeling. Furthermore, DeepTalk-based synthetic speech is shown to be almost indistinguishable from real speech in the context of speaker recognition.