Promise or Peril? Exploring Black Adults' Perspectives on the Use of Artificial Intelligence in Health Contexts
This study is valuable because it moves beyond generic claims about trust or distrust in health AI and shows a more layered picture: cautious optimism, skepticism, and structural critique. Its strongest contribution is the community-centered framing of AI as a possible shield against inequity, while its limits are appropriately bounded by a small, localized qualitative sample and explicitly non-generalizable survey comparisons.
Video Figure
Axes Lens
Rare contribution shape, typical evidence profile. The point here is not a score. It is to show what kind of claim the paper makes, and whether the evidence pattern is unusual or baseline in this 268 -review set.
Contribution shape
- Knowledge form
- descriptive knowledge typical · 92/268
- Novelty type
- empirical finding typical · 68/268
- Abstraction level
- field typical · 41/268
- Generalization target
- user population typical · 75/268
- Validation mode
- mixed methods typical · 136/268
Evidence profile
- Evidence strength
- strong typical · 158/268
- Claim alignment
- strong typical · 231/268
- Overclaim risk
- low typical · 53/268
Review Summary
This paper is a strong CHI contribution because it addresses a real gap in health AI research: the field has often speculated about what marginalized communities might think about AI in healthcare, but here the authors directly ask Black adults and then carefully analyze the resulting perspectives. The study’s main value is not a new algorithm or interface, but a substantive empirical account of how participants reason about AI in relation to structural racism, access barriers, medical mistrust, and the uneven realities of healthcare delivery. The paper is especially compelling when it shows that participants are not simply pro- or anti-AI; instead, they articulate a spectrum that includes enthusiasm for AI as a way to introduce fairness, skepticism about bias, and concern that AI still depends on human institutions that may act unfairly. The “AI as armor” framing is a memorable and useful conceptual contribution because it captures how participants imagine AI as a defensive resource against systemic failure rather than just a productivity tool. The discussion of intended, unintended, and inevitable harms also helps sharpen how researchers talk about AI risk in health contexts. At the same time, the paper is careful about scope: the authors explicitly note that the study is based on 18 Black adults from one Southeastern U.S. metropolitan area, that the sample skews younger and includes some healthcare-experienced participants, and that the survey data were designed to inform focus group topics rather than support statistical generalization. That makes the evidence strong for the studied context but not broadly generalizable. Overall, this is a thoughtful, well-grounded, and field-relevant paper that advances participatory health AI research by centering community voice and by offering concrete conceptual language for future equity-oriented work. Its claims are ambitious but mostly well matched to the evidence because the authors repeatedly frame the contribution as a qualitative characterization of a specific sample rather than a universal account of Black communities.
What Changed
Canon before
Prior dominant assumption in HCI and AI health research assumes AI benefits health primarily through accuracy, efficiency, and precision improvements and that public attitudes towards health AI are broadly negative or uncertain, including in marginalized populations.
Departure from common sense
This paper breaks the typical assumption that Black communities are mostly distrustful or negative about health AI, showing instead a spectrum that includes cautious optimism and a view of AI as potential armor against systemic healthcare bias. It also challenges the notion that AI harms are only unintended by adding a community-framed perspective on intended, unintended, and inevitable harms differentiated by user intention and systemic realities.
Actual novelty
The paper provides the first qualitative study focused specifically on Black adults' nuanced perspectives on health AI and racial health equity, revealing a complex spectrum of views ranging from optimism about AI's potential to improve access and fairness to concerns about biases, mistrust in human providers deploying AI, and structural barriers. It also introduces the framing of AI as 'armor' against systemic healthcare failures and highlights the importance of intention-focused framings of AI harms (intended, unintended, inevitable). It advances participatory health AI research by centering marginalized voices in health AI futuring with community-oriented reflections and proposes new design and research considerations for equity-focused AI.
Evidence
The paper's evidence is grounded in 18 Black adult participants through a mixed-method study combining online surveys and in-person workshops. The authors explicitly note that the survey data were not used for statistical inference, but rather to guide qualitative discussion, while the workshop findings provide the main basis for the nuanced claims about optimism, skepticism, structural barriers, and AI equity concerns.
“ We advance health AI research by articulating previously-unreported health AI perspectives from a population experiencing significant health inequities, and presenting key considerations for future work.”
actual novelty · Abstract · confidence 0.95
“While our survey findings demonstrate that our participants expressed greater enthusiasm for health AI than in the general populations surveyed in prior work, our qualitative findings shed light into the nuanced outlook our participants expressed regarding the use of AI in health contexts. Their feelings ranged from enthusiasm for AI, to skepticism, to feeling that AI is dangerous”
departure from common sense · 6 Focus Group Findings · confidence 0.93
“ Notably, our research was conducted with a racial community–Black adults–from a single metropolitan area within the Southeastern United States, and thus, was designed to provide contextual generalizability”
limitation · 7.5 Limitations and Future Work · confidence 0.98
“As a critical first step to building equitable AI systems, we conducted a mixed-method study to examine Black communities’ current perspectives on health AI. From June-July 2024, we engaged 18 Black adults in a series of in-person workshops and online surveys; all study sessions were held in a city within the Southeastern region of the United States”
validation scope · 3 Methods · confidence 0.96
Limits
Method limits
Study is limited to 18 Black adults from a single metropolitan area in the Southeastern United States, with a sample that skews younger and includes a notable share of participants with healthcare experience. The authors also state that the survey component was used to inform qualitative discussion rather than support statistical conclusions.
Deployment limits
The study does not evaluate deployed AI systems or clinical outcomes; it reports perceptions and future-oriented reflections. Translating these findings into system design, policy, or healthcare practice requires additional participatory work, implementation research, and validation in other settings.
Boundary conditions
Findings are most applicable to Black adults in the studied Southeastern U.S. metropolitan context and may differ across other racial, geographic, cultural, or healthcare contexts. Perspectives may also vary with participants' AI familiarity, healthcare experience, and intersecting identities.
Position in field
Fills a notable gap in HCI and health AI by centering Black adults' perspectives on health AI and racial equity. The paper contributes a community-grounded account of how marginalized users imagine AI as both a possible tool for fairness and a potential amplifier of inequity, strengthening participatory and equity-focused health AI scholarship.