Social media platforms enable diverse users to engage in everyday political talk with (un)known audiences.Platform features and affordances may shape political discussions and how audiences make sense of them, potentially shifting political attitudes.Using affirmative action (AA) -a controversial, identity-centric higher education policy -as a context for analysis, we investigate social media features' and affordances' role in AA discussions.Our qualitative content analysis of over 38,000 social media posts and comments across Reddit, Twitter/X, and TikTok demonstrates how features (e.g., Green Screen) and affordances (e.g., visibility) shape the presentation of external evidence and cues on social media that help users determine information veracity.We introduce sociotechnical rhetorical strategies to describe rhetorical devices enabled by platform features and affordances and consider how these strategies are used to express and refute racism online.Finally, we suggest ways that social media designers may leverage visibility, navigability, and association affordances to enhance users' ability to make sense of and safely experience AA discussions.

The historically controversial U.S. college admissions process is increasingly shaped by algorithmic systems, exacerbating the potential for controversies over admissions and their fairness. Despite their increased use, questions remain about how vendors who provide algorithmic admissions technologies legitimize them and how applicants perceive these technologies. We report on 1) a qualitative content analysis of admissions technology vendor websites, and 2) interviews with college applicants, highlighting the distance between vendors' claimed benefits for universities (e.g., increased decision-making efficiency) and applicants (e.g., ''unbiased'' decisions) and applicants' perceived harms to themselves (e.g., undermining holistic review, hindering diversity, equity, and inclusion efforts). We consider the implications of algorithmic admissions decision-making, including privacy harms, discuss regulatory implications, and offer recommendations to guide algorithmic transparency efforts. However, we caution that transparency would not address some harms perceived by applicants, like inaccuracy and privacy violations.

Emotion AI is increasingly used to automatically evaluate asynchronous hiring interviews. Although touted for increasing hiring fit and reducing bias, it is unclear how job-seekers perceive emotion AI-enabled asynchronous interviews. This gap is striking, given job-seekers' marginalized position in hiring and how job-seekers with marginalized identities may be particularly vulnerable to this technology's potential harms. Addressing this gap, we conducted exploratory interviews with 14 U.S.-based participants with direct, recent experience with emotion AI-enabled asynchronous interviews. While participants acknowledged the asynchronous, virtual modality's potential benefits to employers and job-seekers, they perceived harms to job-seekers associated with automatic emotion inferences that our analysis maps to distributive, procedural, and interactional injustices. We find that social identity can inform job-seekers' perceptions of emotion AI, extending prior work's understandings of the factors contributing to job-seekers' perceptions of AI (broadly) in hiring. Moreover, our results suggest that emotion AI use may reconfigure demands for emotional labor in hiring and that deploying this technology in its current state may unjustly risk harmful outcomes for job-seekers - or, at the very least, perceptions thereof, which shape behaviors and attitudes. Accordingly, we recommend against the present adoption of emotion AI in hiring, identifying opportunities for the design of future asynchronous hiring interview platforms to be meaningfully transparent, contestable, and privacy-preserving. We emphasize that only a subset of perceived harms we surface may be alleviated by these efforts; some injustices may only be resolved by removing emotion AI-enabled features.

Migrants experience unique needs and use social media, in part, to address them. While prior work has primarily focused on migrant populations who are vulnerable socio-economically and legally, less is known about how highly educated migrant populations use social media. Additionally, a growing body of work focuses on algorithmic perceptions and resistance, primarily from laypersons' perspectives rather than people with high degrees of algorithmic literacy. To address these gaps, we draw from interviews with 20 Chinese-born migrant technology professionals. We found that social media played an integral role in helping participants meet their unique needs but that participants perceived social media algorithms to negatively shape the content they consumed, which ultimately influenced their mobility-related aspirations and goals. We discuss how findings challenge the promise of algorithmic literacy and contribute to a human-centered conceptualization of algorithmic mobility as socially and algorithmically produced motion that concerns the movement of physical bodies and interactions as well as associated digital movement. Specifically, we introduce a fourth dimension of algorithmic mobility: algorithmically curated content on social media and elsewhere based on facets of users' identities directly influences users' mobility-related aspirations and goals, such as how, when, and where they go. Finally, we call for transnational policy interventions related to algorithms and highlight design considerations around content moderation, algorithmic user-control, and contestability.

Emotion AI, or AI that claims to infer emotional states from various data sources, is increasingly deployed in myriad contexts, including mental healthcare. While emotion AI is celebrated for its potential to improve care and diagnosis, we know little about the perceptions of data subjects most directly impacted by its integration into mental healthcare. In this paper, we qualitatively analyzed U.S. adults' open-ended survey responses (n = 395) to examine their perceptions of emotion AI use in mental healthcare and its potential impacts on them as data subjects. We identify various perceived impacts of emotion AI use in mental healthcare concerning 1) mental healthcare provisions; 2) data subjects' voices; 3) monitoring data subjects for potential harm; and 4) involved parties' understandings and uses of mental health inferences. Participants' remarks highlight ways emotion AI could address existing challenges data subjects may face by 1) improving mental healthcare assessments, diagnoses, and treatments; 2) facilitating data subjects' mental health information disclosures; 3) identifying potential data subject self-harm or harm posed to others; and 4) increasing involved parties' understanding of mental health. However, participants also described their perceptions of potential negative impacts of emotion AI use on data subjects such as 1) increasing inaccurate and biased assessments, diagnoses, and treatments; 2) reducing or removing data subjects' voices and interactions with providers in mental healthcare processes; 3) inaccurately identifying potential data subject self-harm or harm posed to others with negative implications for wellbeing; and 4) involved parties misusing emotion AI inferences with consequences to (quality) mental healthcare access and data subjects' privacy. We discuss how our findings suggest that emotion AI use in mental healthcare is an insufficient techno-solution that may exacerbate various mental healthcare challenges with implications for potential distributive, procedural, and interactional injustices and potentially disparate impacts on marginalized groups.

Algorithmic systems have infiltrated many aspects of our society, mundane to high-stakes, and can lead to algorithmic harms known as representational and allocative. In this paper, we consider what stigma theory illuminates about mechanisms leading to algorithmic harms in algorithmic assemblages. We apply the four stigma elements (i.e., labeling, stereotyping, separation, status loss/discrimination) outlined in sociological stigma theories to algorithmic assemblages in two contexts : 1) "risk prediction" algorithms in higher education, and 2) suicidal expression and ideation detection on social media. We contribute the novel theoretical conceptualization of algorithmic stigmatization as a sociotechnical mechanism that leads to a unique kind of algorithmic harm: algorithmic stigma. Theorizing algorithmic stigmatization aids in identifying theoretically-driven points of intervention to mitigate and/or repair algorithmic stigma. While prior theorizations reveal how stigma governs socially and spatially, this work illustrates how stigma governs sociotechnically.

Vulnerability intimately shapes the lived human experience and continues to gain attention in computer-supported cooperative work and human-computer interaction scholarship broadly, and in social media studies specifically. Social media comprise sociotechnical affordances that may uniquely shape lived experiences with vulnerability, rendering existing frameworks inadequate for comprehensive examinations of vulnerability as mediated on social media. Through interviews with social media users in the United States (N = 20) and drawing on feminist conceptualizations of vulnerability and social media disclosure and privacy scholarship, we propose a feminist taxonomy of social media vulnerability (FSMV). The FSMV taxonomy reflects vulnerabilitysources, states, andvalences, within which we introduce the state ofnetworked vulnerability andambivalent, desired, andundesired valences. We describe how social media enable forms of vulnerability different from in-person settings, challenge framings that synonymize vulnerability with risk/harm, and facilitate interdisciplinary theory-building. Additionally, we discuss hownetworked, ambivalent, andun/desired vulnerability extend and diverge from prior work to create a theoretically rich taxonomy that is useful for future work on social media and vulnerability. Finally, we discuss implications for design related to granular control over profile, content, and privacy settings, as well as implications for platform accountability, as they pertain to social media vulnerability.

First-generation, low-income (FGLI) students face barriers to college access and retention that reproduce socioeconomic inequities. These students turn to social media for college-related social support. However, while students can reap benefits from social media, it is crucial to investigate under what conditions social media interactions facilitate or hinder students' access to college-related social support. We conducted in-depth, semi-structured interviews with 20 FGLI students in the United States who applied for college in the 2020-2021 application cycle. Our findings illustrate how FGLI identity disclosures on social media can facilitate access to college-related social support when met with supportive or neutral responses, while stigmatizing reactions can disrupt access to these benefits. We draw from the lenses of the "doubly disadvantaged'' and "privileged poor'' used to describe FGLI students in post-secondary education to argue that engaging in FGLI identity disclosures on social media can help students become academically and psychosocially prepared for collegiate environments. Finally, we discuss the implications of this work for theoretical frameworks centering social media and social support, consider when stigma might lead to support space abandonment, and describe the potential implications for social media design.

Pregnancy loss is a common yet stigmatized experience. We investigate (non)disclosure of pregnancy loss among LGBTQ people to known ties on identified social media as well as what constitutes ideal socio-technical disclosure environments. LGBTQ persons experiencing loss face intersectional stigma for holding a marginalized sexual and/or gender identity and experiencing pregnancy loss. We interviewed 17 LGBTQ people in the U.S. who used social media and had recently experienced pregnancy loss. We demonstrate how the Disclosure Decision-Making (DDM) framework explains LGBTQ pregnancy loss (non)disclosure decisions, thereby asserting the framework's ability to explain (non)disclosure decisions for those facing intersectional stigma. We illustrate how one's LGBTQ identity shapes (non)disclosure decisions of loss. We argue that social media platforms can better facilitate disclosures about silenced topics by enabling selective disclosure, enabling proxy content moderation, providing education about silenced experiences, and prioritizing such disclosures in news feeds. CAUTION: This paper includes quotes about pregnancy loss.