“It just feels morally not right to Sell the data”: Ethical and social perspectives on human genomic data sharing in Uganda—A phenomenological qualitative study

Abstract

While genomic data sharing enhances transparency and research efficiency, it also raises significant ethical and social challenges. This study explored stakeholders’ perspectives on these issues, particularly around privacy, confidentiality, and equity in collaborative research. A phenomenological qualitative study was conducted between August and December 2023 at Makerere University College of Health Sciences, other research-intensive institutions, and national regulatory bodies. The study engaged 86 participants: 47 key informants (16 researchers, 14 ethics committee members, nine community advisory board members, and eight research regulators) and four deliberative focus group discussions with 39 participants. Interviews were transcribed verbatim, and thematic analysis was conducted using NVivo 14. Three major themes emerged: (1) stakeholders’ experiences in genomic research, including their roles as participants, implementers, or overseers; (2) ethical concerns, such as informed consent, third-party data access, inequities between high-income and low- and middle-income country (LMIC) researchers and participants, and the lack of benefit-sharing frameworks; and (3) social implications, including stigma, discrimination, labeling, community perceptions of fairness, and the need for meaningful engagement. Participants emphasized the importance of protecting participant rights, promoting equity, and ensuring robust data governance and security. The theoretical frameworks of principlism and distributive justice provided a valuable lens for examining these concerns, particularly by highlighting the need to safeguard privacy and fairly distribute responsibilities and benefits in global collaborations. Participants also noted that perceptions of fairness are shaped by trust, local context, and past experiences with research factors that are critical for building equitable and respectful partnerships. This study underscores the urgent need to strengthen protections for research participants and promote fairness in genomic data sharing. Policies should, if adopted, emphasize culturally contextualized consent, active community engagement, restricted third-party data access, and strong data protection mechanisms to address existing inequities and prevent misuse.

Introduction

Advancements in sequencing combined with artificial intelligence (AI) have significantly accelerated genetics and genomics research (Mulder et al., 2017; Paul et al., 2024; Xu et al., 2019). AI is becoming a bigger part of genomics helping scientists analyze the data and make new discoveries to improve research and patients care. The ability of AI technologies and tools to process large amounts of data, identify patterns, and make predictions has not only accelerated research but also enhanced the accuracy and efficiency of genomic studies (Lin and Ngiam, 2023). Machine learning is a key tool that powers algorithms to automatically learn from data and enabling predictions or decisions (Lin and Ngiam, 2023). Machine learning is increasingly being utilized in analyzing genomic data and this helps to uncover patterns, associations, and insights that are not readily noticeable through traditional methods. By applying these methods to genomic data, researchers can enhance the understanding of complex biological systems, improve diagnostics, and contribute to precision medicine.

Genomics research heavily relies on the sharing of data via global collaborative networks such as the Genome Consortium, the Global Alliance for Genomic Health (GA4GH), Clinical Genomic Database, Human Genome Database, International Cancer Genome Consortium (ICGC), and the Global Alliance for Genomic Health (GA4GH; North, 2015; Rehm et al., 2021; Zhang et al., 2019), European Genome-Phenome Archive (EGA; Freeberg et al., 2022), Clinical Genomic Database, data base of Genotype and phenotype (dbGaP; Mailman et al., 2007), Human Genome Database (Bruford et al., 2008), H3Africa Bionet (Beiswanger et al., 2017). Sharing genomic data through global collaborative networks offers numerous potential benefits, including advancements in personalized medicine, cost-effectiveness, enhanced transparency, increased diversity in research, and contributing to the common good (Horton and Lucassen, 2023).

Despite the benefits of sharing genomic data, there is a need to consider the ethical and social implications particularly in low-and-middle-income countries (LMICs). In these regions, genomic research, although experiencing rapid growth, remains in the infancy stage (Omotoso et al., 2022; Rahimzadeh et al., 2020). This developmental lag can be attributed to several factors, including inadequate infrastructure, limited resources, a pressing need for enhanced capacity building among researchers, and the absence of comprehensive ethico-regulatory frameworks, all of which hinder progress in collaborative human genomic research (Chatfield et al., 2021; Helmy et al., 2016; Mboowa et al., 2021). The act of sharing genomic data introduces significant ethical and social challenges, such as the risks associated with re-identification, the commercialization of shared data, inequitable distribution of benefits, ethics dumping due power dynamics between high income and LMICs (Chatfield et al., 2021), and perceptions of unfairness and inequality among researchers from LMICs (Chennells and Steenkamp, 2018; Goodman et al., 2017; Kaawa-Mafigiri et al., 2023; Munung et al., 2017). Given the uniquely personal nature of genomic data, the challenge of achieving complete anonymization or de-identification remains persistent, despite ongoing efforts to mitigate these issues as well as concerns in the consenting process (Otlowski and Eckstein, 2023; Thomson and Peabody, 2022; Tindana et al., 2020) The potential for unauthorized access or misuse of genomic data exacerbates the risk of privacy breaches, even when robust safeguards are in place, thereby leading to issues of trust (Jackson et al., 2019). Socially, these risks carry profound consequences such as stigma, discrimination, labeling, and shame, adversely affecting mental health which may be as a result of risk of re-identification and loss of privacy.

The Uganda Medical Informatics Centre (UMIC) is currently one of the largest health research-orientated computational resources in Sub-Saharan Africa with modern high-performance computing facilities (Fatumo et al., 2022). Genomic research in Uganda has experienced significant growth over the past decade, positioning the country as one of the leading hubs for genomic science in Africa. This advancement is driven by Uganda’s rich genetic diversity, robust institutional infrastructure, and active participation in international collaborations (Mboowa et al., 2021). In Uganda, bioinformatics was initially mainly done by individual groups at Makerere University and the Uganda Virus Research Institute (UVRI). H3 Africa https://h3africa.org/ birthed four genomics projects at Makerere University. These included the Collaborative African Genomics Network (CAfGEN), Integrated Biorepository of H3Africa Uganda (IBRH3AU), Integrated approach to the identification of genetic determinants of susceptibility to trypanosomiasis (TrypanoGen), and Nurturing Genomics and Bioinformatics Research Capacity (Mboowa et al., 2021).

The sharing of genomic data raises significant ethical and social concerns that must be addressed to ensure the responsible use of such sensitive information. The risk of data misuse, potential for non-research uses and commercialization without a benefit sharing framework present ethical concerns. This study explored the ethical and social considerations that influence data-sharing practices in collaborative human genomic research. It specifically examined how these practices affect the rights of research participants and researchers from LMICs, who often face power imbalances in such collaborations. The findings of this research may identify strategies to protect the rights of research participants and researchers from LMICs involved in collaborative research. Ultimately, this study aims to contribute to the development of guidelines and policies which if adopted may promote fairness, equity and responsible data sharing.

Materials and methods

Study design

We used a phenomenological qualitative research design (Cypress, 2018) that explored the told experiences of key stakeholders in research including genomic researchers, research ethics committee (REC) members, national regulators, and members of community advisory boards (CABs) through key informant interviews. We also conducted deliberative focus group discussions with individuals that had participated in genomic research in the previous 3 years. The study combined focus group discussions (dFGDs) and key informant interviews (KIIs) to gather rich, nuanced data from diverse perspectives. This mixed-methods approach facilitated a comprehensive understanding of the research topic by triangulating data sources, thereby enhancing trustworthiness of the findings.

Theoretical perspective

This study is grounded in two ethical frameworks, principlism by Beauchamp and colleagues (Beauchamp, 2016; Beauchamp and Childress, 2019), and Rawls’ distributive justice theory (Rawls, 1971). Principlism serves as the overarching framework and is built around four core principles, respect for autonomy, non-maleficence, beneficence, and justice. Rawls’ theory of distributive justice (Rawls, 1971) introduces the idea of justice as fairness, justice involves the fair allocation of resources, rights, and opportunities in a manner that respects each person’s equal claim to basic liberties.

In this study, the main ethical issues that arise revolve around informed consent, ownership and control of genomic data, equitable access and benefit-sharing, and concerns about ethical imperialism. Each of these issues connects back to the guiding ethical principles. The issues of informed consent relate to autonomy, while fears of data misuse and the social risks are tied to non-maleficence. The justice-related concerns, such as unequal benefit-sharing or fears of ethical imperialism, are addressed through Rawls’ distributive justice framework. To emphasize fairness, we invoke Rawls’ veil of ignorance principle (Rawls, 1971), a concept that challenges us to design systems without knowing our own position in society, encouraging decisions that are fair and impartial for everyone.

Setting

The study was conducted at Makerere University College of Health Sciences and its affiliate research intensive institutes. Additionally, we engaged national regulatory entities, including the Uganda National Council for Science and Technology (UNCST). We also included members of the seven accredited Research Ethics Committees (RECs) that review genomic protocols in Uganda, and members of community advisory boards (CABs) affiliated to the study sites.

Sampling

We used a purposive sampling technique and the maximum variation principle (Berndt, 2020; Campbell et al., 2020) to recruit a diverse range of stakeholders based on their role in genomic research or oversight in Uganda. We aimed to capture their varying views on data sharing in collaborative genomic research. We identified potential key informants via the National Research Information Management System (NRMIS), a centralized digital platform for coordinating and tracking research activities nationally and our professional contacts at relevant bodies.

All potential key informant participants were contacted directly either via email or telephone, while genomic research participants were contacted only via telephone. They were all individually provided with information about the study, and enrolled after providing informed consent.

Selection bias was avoided by clearly ensuring the criteria for selecting participants aligned closely with the study’s purpose and objectives as well as setting clear boundaries for inclusion and exclusion to avoid favoring specific categories.

Participants

A total of 86 research stakeholders participated in this study, comprising 16 researchers, 14 members of RECs, nine members of CABs, eight members of regulatory bodies, and 39 participants involved in genomic research. The researchers included principal investigators and study coordinators who were engaged in studies with a genetics or genomic focus. The selection of CAB members was intentional, as they serve to represent community perspectives and are continuously trained to communicate research information in an accessible manner to community members.

Research team and reflexivity

The data collection team comprised three members, led by DES, and included two trained research assistants. To minimize bias, interviews were conducted by team members who had no supervisory or power over participants. The team was trained and experienced in the conduct of qualitative interviews.

Recognizing the importance of neutrality, especially when interviewing research stakeholders, the team set aside personal perspectives. During the study and reporting process, the researchers-maintained objectivity by actively listening and focusing on the interviewee’s responses without forming personal opinions or judgments during the conversation despite prior involvement in the conduct of genomic collaborative research. We recognized that our roles as researchers with professional experience in genomics and research ethics could shape how we interpreted participant responses. To mitigate potential bias, interviewers followed a standardized topic guide and were trained to use neutral probing techniques to avoid leading participants. The team also engaged in regular reflexive discussions during team meetings to critically reflect on our assumptions and how these might influence data collection and analysis. We were aware that the pre-discussion briefings on genomic research that were intended to promote understanding, may have unintentionally shaped responses during focus group discussions. We set several steps to manage this. The briefings were designed to be neutral, fact-based, and free of leading language and the facilitators were trained to avoid reinforcing specific viewpoints and to encourage a range of perspectives. We were also aware that in the focus groups, some voices are louder than others. To address these issues, we maintained a neutral tone, established clear ground rules and emphasized the importance of equitable participation at the outset of each session as well as used open-ended questions, and encouraged independent views. This approach aimed to ensure that all participants had the opportunity to contribute their perspectives. Data analysis and interpretation were collaboratively carried out by DES, JB, and AS over the course of the study. During analysis, coding was conducted independently by multiple team members, followed by consensus meetings to ensure that emerging themes reflected participants’ perspectives rather than our own.

Research instruments

We selected two instruments together to provide both depth (through detailed interviews) and breadth (through group discussions), allowing the team to explore complex issues more thoroughly. The schedule for each instrument was developed by first defining the objectives of both instruments and clarifying the specific information we aimed to gather from the interviews and focus group discussions. We then outlined the main topics and themes that needed to be addressed by each method.

The focus group and interview guides were developed from the literature (Chawinga and Zinn, 2019; Creswell and John, 2018; Goodman et al., 2017; Horton and Lucassen, 2023; Rahimzadeh et al., 2020), and subsequently revised to capture new emerging questions. The discussion guide included questions related to awareness of data sharing in genomics, attitudes toward data sharing, and ethical and social concerns regarding sharing genomic data. The guides were professionally translated into Luganda in consultation with community members to ensure culturally appropriate meanings of genomic terms. Both guides featured open-ended questions, and were piloted and revised before the full data collection began. Data were collected until no new insights emerged regarding ethical and social concerns.

Data collection

A list of individuals who had participated in genomic research either currently or within the past 3 years—was compiled from various studies conducted across different institutions. Eligible individuals were contacted by phone or email and invited to participate in the study. Data were collected between August 2023 and March 2024

Written informed consent was obtained before prior to enrollment in the study. The interviews and focus groups were conducted in English or Luganda, the most commonly spoken local language, based on the respondents’ preference. All interviews and focus groups were audio recorded with the permission of the participants. No personal identifying information was collected to protect confidentiality.

Key informant interviews

We conducted interviews both face-to-face and virtually based on respondent availability, preference, and convenience.

Participants were encouraged to choose a time and location that ensured their privacy, and they were informed in advance about the nature of the call to help them prepare accordingly. Interviews lasted approximately 40–60 minutes each, and were conducted in both English and Luganda, the most widely spoken language in Central Uganda, to accommodate participant preferences. The interview guide has been submitted as a Supplemental File.

Deliberative focus group discussions

The focus groups were conducted in Luganda. They were held in private spaces within the participating organizations with each lasting approximately 60–120 minutes.. Given the complexity and relative novelty of genomic research in Uganda, the sessions began with a neutral 20-minute briefing led by DES. This briefing covered basic genomic concepts such as the definition of genes, the benefits and risks of genomic research, and the rationale for data sharing, including the use of international repositories. The aim was not to influence responses but to ensure participants could engage meaningfully, especially since some genomic terms lacked direct equivalents in Luganda. Participants were encouraged to ask questions, and any misconceptions were addressed. Before the discussion, participants completed a short questionnaire capturing demographic details and prior exposure to genomic research to help guide facilitation. A 15-minute break followed the briefing, after which the focus group discussions commenced. We conducted a total of four focus groups involving 39 individuals, including young adults, their parents or guardians, and a mix of older and younger adults. Each group included both male and female participants. The discussions provided rich insights from across different age groups and backgrounds, helping us build a well-rounded understanding of the ethical, social, and practical issues surrounding genomic data sharing. Participants were asked open-ended questions such as, “What are the benefits of sharing genomic data?” (see appendix for full guide) to encourage open and thoughtful discussion. Their responses helped us explore not just opinions, but also the values and concerns shaping those views. Each session was carefully documented through note-taking and audio recordings to capture the conversations accurately. To maintain the quality of the data, the research team held debriefing meetings after every session to review completeness and reflect on emerging themes.

Data management and analysis

Audio recordings were transcribed verbatim, verified for accuracy, and de-identified to maintain confidentiality. The lead researcher verified the transcripts to ensure their accuracy and authenticity with respect to the audio recordings.

Thematic analysis was utilized following Braun and Clarke’s framework to identify, analyze, and interpret data (Clarke and Braun, 2017). A team-based approach (Creswell and John, 2018) was adopted for data analysis, involving DES, JB, and AS, who developed a codebook through multiple readings of the data. This process included marking significant content and taking notes, leading to synthesized codes and refined themes through team discussions until consensus was achieved (Creswell and John, 2018; Nabukenya et al., 2023). The analysis was systematic and iterative, allowing for continuous theme refinement as new data emerged. Initially, a deductive approach was used with a draft framework from early interviews, while also remaining open to inductive insights. Three final themes emerged from the data including experiences, ethical concerns, and social issues. These themes were derived from both data sources, with key informant interviews placing particular emphasis on ethical concerns related to AI. NVivo software facilitated data management, and the COREQ checklist was used to guide the presentation of findings (O’Brien et al., 2014).

Trustworthiness

Triangulation of data sources was also done as a key strategy to enhance trustworthiness. By integrating findings from focus groups and interviews, the study was able to present a more comprehensive and nuanced understanding of the research topic. To enhance the trustworthiness of the study, several strategies were employed (Gonella et al., 2021; Lincoln, 1995). To reduce threats to validity, the study employed a qualified team that underwent protocol training. Development of tools was done and consensus about differences in meaning were reached following discussion as a team. Translation of the tools was done by an independent reviewer conversant in both languages who reviewed them. Transcripts were reviewed by the lead researcher to ensure they are a true reflection of the recordings. Triangulation of data sources was done by examining evidence from the different sources and using it to build justification for themes. Debriefing sessions were conducted with colleagues to discuss the interviews and discussions, the analysis process and findings.

Results

Demographics

A total of 86 stakeholders participated in this study, contributing valuable insights into the ethical and social dimensions of genomic data sharing. In total, the study engaged 34 female and 52 male participants. Of these, 47 individuals took part in the interviews. This group included 16 researchers, 14 REC members, 8 regulators, and 9 CAB members.

The remaining 39 participants were involved in dFGDs, which included 17 female and 22 male participants from diverse backgrounds (Table 1). This mix of key informants and focus group participants allowed for both depth and breadth in exploring stakeholder perspectives on genomic data sharing.

Table 1.

Summary of the dFGD participants age and gender distribution.

dFGD	Participant description	Male participants	Female participants	Total participants
dFGD 1	Mixed group	6	5	11
dFGD 2	Young adults	5	5	10
dFGD 3	Parents and guardians	6	4	10
dFGD 4	Mixed group	5	3	8
Total		22	17	39

Thematic analysis of the KIIs and dFGDs revealed three main themes, which are summarized in Table 2. These themes include: experiences in the conduct of genomic research, ethical considerations, and social considerations.

Table 2.

Emergent themes on the ethical and social concerns around genomic data sharing.

No.	Theme	Category
	Experiences in the conduct of genomic research	Current participation in genomic studies Previous experience in genomic studies Voice of the community Research oversight experience
	Ethical concerns	Informed consent Unlimited access of data by third party companies Commercialization No framework for benefit sharing Perceived risk of AI (Risk of de-identifiability, breach of confidentiality and privacy Power differentials Fear of data misuse
	Social concerns	Socio-cultural views around sharing genomic data Individual and community stigmatization Labeling Community engagement Community perception of data sharing Shame

Theme 1: Experiences in the conduct of genomic research

All participants shared firsthand experiences related to genomics research. Genomics research participants, whether current or former, demonstrated an awareness of their involvement in such studies. Many associated genomics research with concepts like paternity and DNA testing. Several participants seemed surprised that their data would be shared globally.

While several of the participants indicated that data was valuable to them, a few expressed that they didn’t attach value to the shared data thus:

I value this data very much because it is from me and if misused, it can affect me, my children and my parents (dFGD 02)

Personally, that data is not valuable to me—not so much because I don’t know anything. So, I don’t really see value in it (dFGD 03).

Furthermore, REC members reflected on their experiences approving genomics studies, highlighting challenges in navigating the complexities of the difficult terminology during the review process. CAB members underscored their role as the voice of their communities, emphasizing their responsibility to represent community interests and serve as a crucial link between researchers and the communities. National regulators shared their experiences of providing oversight for genomics studies, reflecting on the processes and responsibilities associated with evaluating submitted research proposals.

Theme 2: Ethical concerns in the sharing of genomic data

The ethical concerns described included informed consent, unlimited access of data to third parties, power dynamics, and inequity and fairness

Informed consent.

Reflections from the findings show that the informed consent process in genomic research must go beyond practical formality to reflect culturally sensitive and participant-centered communication. The participants stressed that consent processes should explicitly address data sharing, and participants should be provided with all the information including how genomic data may be used by third parties. Participants’ preferences and concerns should guide the design of consent materials. Several participants noted that current practices often fall short due to the technical complexity of genomic terminology, which can hinder understanding for participants and undermine the principle of voluntary, informed participation.

Stakeholders also identified systemic limitations, such as the limited capacity of ethics review bodies and researchers to develop and assess consent forms that are transparent, comprehensible, and aligned with community values. This lack of clarity, they argued, can erode trust and lead to ethical challenges when participants later feel misled or uninformed about how their data are used.

One CAB member indicated that the research participants involved in genomic studies may not fully understand what the study is about, let alone being aware that their data will be shared with other scientists. He attributed this to the complex language used in consent documents.

I am the voice of the community. You know, the researchers want to bring their consent forms in their difficult language and then they say, we are researching about this and that after we shall present the findings to others forgetting the owners of the data. These participants have not studied much and may not understand. The researchers say write here, sign here,” and they end up signing what they don’t understand. So, we have to interpret for them (CAB member 01).

Unlimited access to data by third parties.

Reflections from the data highlight a deep and ongoing debate among researchers about who should have access to genomic data especially when it involves third parties like funders, collaborators, or institutions outside the country where the data was collected. On one hand, some researchers were clearly uncomfortable with the idea of unrestricted third-party access, particularly when it was imposed by external funders. They felt that being required to hand over data without much say in how it’s used or by whom was not just inconvenient. They described unrestricted data access particularly when mandated by external funders as a form of ethical imperialism where more powerful institutions often from high-income countries set the rules, leaving local researchers with little choice but to comply if they wanted to receive funding. In their view, such practices compromised local ownership and autonomy, as researchers were often required to accept broad data sharing conditions to secure funding. These participants argued that this dynamic reflected an imbalance of power in global research collaborations, where institutions from high-income countries set the terms with limited negotiation.

However, other researchers did not frame the issue in terms of ethical imperialism. Instead, they emphasized the scientific benefits of broad data sharing and felt that, if appropriate safeguards were in place, third-party access could enhance collaboration and accelerate discovery. This divergence in opinion illustrates the tension between promoting open science and ensuring equitable partnerships in genomic research.

So, there are very big issues with what I will call ethical imperialism where by the donors for example we are now having grants which are only given if you sign to unrestricted access by third parties; both the specimens, Bio specimens and associated data. And so, a grant comes in like that and you ask yourself is this ethical (REC member 02).

Furthermore, researchers expressed concerns about the risks of unrestricted sharing, including potential misuse, loss of privacy, and challenges in protecting participants’ rights. Those with divergent views on imperialism had no reservations as long as there are clear data access guidelines in place and they cited advancement of science, transparency, efficiency, and local capacity strengthening as the main reason for their choice. This divergence in opinion illustrates the tension between promoting open science and ensuring equitable partnerships in genomic research.

It’s a bit tricky. There are valid arguments, both for and against. Arguments for of course, is to really advance research and science, make it quicker without necessarily reinventing the wheel, it cuts on expenses. But also, I think for the investments some people have put in you; it’s like you’re giving back because this money is not ours it’s like for a common good (Researcher:12)

Some of the participants particularly CAB members equated data sharing to a form of colonialism and noted:

R: I think it is taking us back to the new rebirth, colonialism and those practices that were done in World War 1. What happens when you infect people with syphilis and you don’t treat it? Because someone has all the determinants of your genome and the rest of the studies, can be lab studies. Just from the data, samples, you don’t have to come back to Africa for some of those studies. For me, it is scary. And that has been happening where someone looks and sees that this was done here and this was done from another place (CAB member 08).

Data sharing with commercial entities and benefit sharing.

Participants strongly felt that researchers and communities should equitably benefit from the outcomes of genomic data sharing, particularly when commercial entities are involved. However, several noted that in practice, communities often do not share in these benefits due to weak or nonexistent regulatory and benefit-sharing frameworks. Concerns were raised about how sequence data is frequently used to develop diagnostic tools and therapeutic products without any tangible return to the individuals or communities that contributed the samples. One researcher attributed this disconnect to a moral gap highlighting how the lack of transparency and, in some cases, dishonesty in how benefit-sharing arrangements are communicated and implemented.

And I don’t know if whoever has made a product is transparent enough to declare how much they’ve made up of it, to ensure that there’s a bit of financial benefit to those who have contributed their time and samples to the study. Although, remember in the consulting process, this is a voluntary activity so, there’s also I don’t know what to call it, contradiction, yes, there’s also that now, contradiction. So technically it’s a bit difficult, but it just feels morally not right to sell the data (Researcher 15).

Several pertinent concerns were raised including the risk of re-identification and breach of confidentiality, and data misuse. A critical concern raised by participants was how to effectively protect and control access to genomic data. The study revealed growing unease about the risk of re-identification, particularly as artificial intelligence (AI) algorithms become more advanced and capable of linking datasets in ways that could compromise participant anonymity.

Risk of re-identifiability:

Participants highlighted the significant risk of re-identifiability of shared genomic data, even when de-identified. This risk arises due to the unique nature of genomic data, which AI can exploit to trace data back to individuals. The concern is amplified in large datasets, where familial linkages can be deduced without laboratory testing. While some participants proposed anonymization as a potential solution to mitigate re-identification risks, others pointed out its significant limitations. Although anonymization can help safeguard individual privacy by preventing direct identification, it can also reduce the utility of the data, especially when researchers need to recontact participants or return clinically significant individual findings. This tension underscores the ethical dilemma in balancing privacy with potential health benefits. Participants emphasized that ethical data handling is not just a procedural requirement but a moral obligation, calling for the adoption of responsible and context-sensitive data governance practices.

With AI (Artificial Intelligence), anything is possible. Because AI can use any kind of data and it can be traced to an individual even where you deidentify. That is why there is a great risk and you need to involve the community leaders and so many influencers at that level. But those risks are real (Researcher 07).

On the other hand, one researcher indicated that the risk of re-identifiability is often low because genomic data is usually not shared with the meta data.

I don’t think that is really possible because it is usually delinked data so, they don’t have the identifying information attached to it. I mean, from my experience, what we have deposited is usually targeted exome sequencing. At the moment I don’t think that is broadly possible, but maybe in the future. Once we have widespread sequencing of the community and the databases, and the sequences for identifiable sequences are available on the internet, you can then sort of match those sequences to those public databases. But I don’t think it is a big risk now (Researcher 8).

Breach of confidentiality and loss of privacy.

The findings revealed that the risk of re-identifiability presents a significant threat to participant confidentiality and privacy. Many stakeholders, particularly research participants, identified the potential loss of privacy as the most serious risk associated with genomic data sharing. They expressed that breaches of confidentiality could negatively impact a person’s self-esteem, confidence, and overall sense of dignity. These concerns were amplified by the large volumes of genomic data collected and the indefinite nature of data storage.

Participants raised critical questions about whether it is realistically possible to safeguard privacy in the context of genomic data sharing. While perspectives varied, research participants in particular voiced feelings of vulnerability and concern. They noted that their decision to participate in genomic studies was grounded in the trust they had in their healthcare providers, and the possibility of data misuse or re-identification felt like a violation of that trust. For them, the risk of privacy was considered personal and deeply tied to their dignity and autonomy.

It may also make me fail to trust health workers generally because of that mistake that may have happened. And I may say that health workers are complicated people; they take our information and they share it or it may look like that they are getting money from our information and they do not care about me as a person. And then I choose to avoid all the medical workers because of what is circulating around (dFGD 04).

Some researchers played down the risk of breach of confidentiality and they related it to a hypothetical situation because in their opinion no one could use one’s DNA.

The people’s data will get known by others but that risk also is more of a theoretical risk than a real risk because you can’t do anything with somebody’s DNA really. There is nothing you can do about it. The interpretation of a genomic-wide association result is not something that many people can make heads and tales. It’s not like this DNA testing that people do in town that says or this is your dad or this is not your dad (Researcher:02).

One community representative indicated that loss of privacy was like giving away one’s identity

Genomics data is really the person’s identity, the map of this person’s life, reduced to perhaps in a test tube or a computer script, which is very scary. And this information does not have time bounds that you can use it today and discard it, or that it will decay or whatever. Once it has gone out, you know, it can be studied over and over and over, even new technologies that come next year, new inventions into the other year, will continue to make more meaning of this person’s information, this person’s basically identity. And because of that, it makes it very risky, it’s the ultimate identity of the individual, it’s like the person giving himself away (REC member 04).

Data misuse.

The findings revealed growing concerns about the potential misuse of genomic data, especially in light of advancing technologies. Several participants expressed fears that shared genomic data could be accessed by unauthorized individuals or institutions, leading to uses beyond the original scope of consent. They emphasized that while de-identification of data is a common safeguard, it may no longer be sufficient. With the capabilities of AI, it is now possible to extract new and potentially sensitive insights from large datasets, even without direct identifiers raising serious ethical questions about privacy and control.

Participants also pointed to the risk of data being repurposed or shared with groups not originally approved or foreseen. In response to these concerns, many stressed the urgent need to strengthen institutional capacity, particularly through the establishment of robust audit trails. Such mechanisms would help ensure transparency and accountability in how genomic data is managed, accessed, and used over time.

The researchers who shared the data have no control over how it is used.

How do you restrain the researchers from limiting their objectives to the questions that they have asked? It becomes a problem. How do we ensure that they don’t use this information for questions that are not asked, questions that are not yet asked? How do we ensure that we control the use of this data beyond the immediate? It becomes a big problem once the data goes out to the different partners (REC member 06).

Theme 3: Social concerns around the sharing genomic data

Participants highlighted a range of social concerns tied to genomic data sharing, including cultural sensitivities, stigma, labeling, mental health issues like depression and community perception of fairness. They attributed these outcomes largely to breaches of confidentiality and misuse of data, underscoring the serious social risks involved. The reflections highlighted the fact that the social concerns brought about from the sharing of genomic data would negatively affect their lives.

Research participants raised several social issues associated with a breach of confidentiality which could potentially reveal their HIV positive status

Because when you are depressed or if the whole village knows that I have HIV, everyone will run away from me. I will be isolated from the community and the more you stay alone, the more you overthink. The more you overthink, you may even decide to leave the ARVs and say, let me die! No one in the village likes me because I have HIV (dFGD 02).

The loss of privacy and confidentiality, leads to harm, it could be political harm, social harm, group harm that arises out of an individual’s choice to participate on behalf of people that have not been consented. It brings a lot of social, cultural political and legal issues. People who have nothing to do with the research are affected that way (REC 02).

Social cultural views around genomic data sharing.

Some participants raised socio-cultural concerns about genomic research, particularly regarding the collection of biological samples such as hair, nails, and saliva. They pointed out that some of these samples are sacred in certain communities and collecting is associated with witchcraft. Additionally, there was uneasiness about sharing of the data because of the potential revelation of sensitive familial relationships, such as paternity, which participants linked to genomic studies. These views highlight the need for culturally sensitive approaches and robust community engagement to address misconceptions and build trust in genomic research.

Traditionally, the members in the community have a negative perception about research and the reason for that they have superstitions about hair, nails and the like and perceive them as accessories for witchcraft. So, what about sharing genetic data? The researchers have to help the community first get reed of cultural bias that people may have (CAB 2)

Some participants were particularly worried about potential findings from the analysis of their genomic data

If it happens that my genes will be associated with certain risks of the conditions, how are they going to protect me? The psychological issues that come with that. And especially now when they have told you and maybe you see trends in your family and with this common kind of diseases that are coming, So, psychological torture in terms of ethical issues, needs to be addressed very much. Are there facilities that offer care in case somebody now says, if I develop this, where do I go? (dFGD 04)

Community perception of fairness

Reflections from the data showed that participants in this study described fairness not only in terms of how benefits are shared, but also in relation to trust, confidentiality, and emotional wellbeing. A strong expectation was placed on the trusted health professionals and researchers. This highlights how participants saw fairness not just in procedural terms (such as getting consent or sharing benefits), but also in relational and emotional terms: being treated with respect, having their dignity upheld, and being protected from harm. Fairness, therefore, involves ensuring that research participation does not expose individuals to psychological harm, community rejection, or reputational damage.

Additionally, across the FGDs, participants strongly expressed the view that fairness in genomic research should include equitable benefit sharing, particularly when their biological materials and data contribute to profitable medical advancements. Many believed that since they voluntarily provided samples and contributed to achieving the research objectives, they should receive a portion of the benefits, whether in the form of financial compensation, access to developed treatments, or recognition.

Others emphasized that any medications, vaccines, or treatments developed as a result of their contributions should not be sold to them

They should not charge us for any vaccine that is developed from the findings because the data is ours… We volunteered and sacrificed our time to be here.” . . … some of us earn more than the transport refund we are given. So, it has to be free of charge because it is an experiment (dFGD 3).

Community engagement.

Insights from the data underscored the critical role of community engagement in genomic research, especially in the context of data sharing. Several participants emphasized that genomic data is inherently sensitive and personal, often carrying implications not only for the individual but also for their families and broader communities. Because of this, participants called for communities to be meaningfully involved in all phases of research from study design to the dissemination of findings. This is more than procedural inclusion; it speaks to the essence of community engagement as a process built on mutual respect, shared understanding, and ongoing dialog. Crucially, the reflections pointed to the early involvement of community leaders as essential. These leaders serve as gatekeepers and trusted representatives, helping to ensure that research is culturally appropriate, ethically sound, and well-received. Their participation reinforces the idea that community engagement is not a one-off activity but a continuous partnership that shapes the relevance, quality, and impact of research.

One CAB member highlighted the importance of community engagement indicating how it should be done from the beginning of the study

There is no way you can come from nowhere and say; we want to tell you about what we discovered around our study from your genomic data without involving them. It is important to involve the community right from the study design while you tell them the reasons why you want to conduct the study and respond to some of their concerns. So, involving them right from the beginning through study implementation and then there’s that area of sharing the findings, to me that solves a number of questions and problems (CAB 04).

Discussion

Our findings highlight key ethical and social issues that should be considered during the sharing of genomic data.

Ethical concerns included challenges to the informed consent process, perceived risks associated with AI, inequitable benefit-sharing, and power imbalances in research collaborations. Social concerns relate to fears of stigma and discrimination and the importance of community engagement. The findings align with the theory of principlism (Beauchamp, 2016; Beauchamp and Childress, 2019), particularly the principles of autonomy, non-maleficence, and beneficence, as well as with Rawls’ theory of distributive justice Rawls (1971), especially in relation to unequal benefit sharing, and fears of ethical imperialism in genomic data use.

The findings suggest that research participants did not seem to be informed about the sharing of their genomic data. Across all stakeholder groups, there was general agreement that informed consent is essential to the ethical sharing of such data. However, some research participants and CAB members appeared unaware of what genomic data sharing actually entails, highlighting gaps in understanding at the community level. Researchers and regulators further noted the limited capacity of RECs to adequately review and assess genomic research protocols, particularly with regard to consent processes. While informed consent is a core ethical requirement in research, it presents significant challenges when applied to genomic data sharing especially in LMICs. These challenges are amplified by the complexity of genomic science and the limited literacy of some participants, which can hinder full understanding of the study and its implications. When participants are not clearly informed about how their data will be used, who will access it, or the potential risks involved, it raises concerns about fairness, justice, and diminished autonomy. Stakeholders such as the REC members, CAB members, participants, and regulators challenged researchers to develop comprehensive, culturally sensitive consent processes that emphasize transparency about procedures and associated risks in genomic data sharing. Studies, including Bukini et al. (2020) and a multi-country study in Ghana, Uganda, and Zambia, emphasize the importance of continuous and simplified consent approaches to enhance understanding. As Selita (2019) notes, for consent to be truly informed, all potential uses of the data must be disclosed something that is often overlooked (Selita, 2019). According to the principle of autonomy (Beauchamp, 2016; Beauchamp and Childress, 2019), participants have the right to receive clear and complete information about the study before making a decision. Their autonomy should be respected not just in the act of giving consent but also in ensuring their privacy and confidentiality are protected. These challenges are even more complex in genomic research, where the science itself can be difficult to explain in simple terms, making meaningful consent even harder to achieve.

Our findings suggest that stakeholders are concerned about the increasing use of AI in genomic research which further complicates the ethical and socio-cultural concerns. While AI has the potential to transform data analysis and interpretation, it also raises concerns about the risk of re-identifiability, confidentiality breaches, and data misuse. Through AI technologies, customized individual genetic profiling can be generated by users in different contexts (Wang and Liang, 2024). Genomic participants across all categories expressed concerns and anxiety about unrestricted access to their data by third-party companies. Even when data is anonymized, participants may still be vulnerable to re-identification or unintended consequences, such as discrimination. The data supports earlier findings by Rahimzadeh et al. (2020) and Selita (2019), which highlight the ethical, social, and legal concerns associated with shared genomic data, particularly the risk of re-identifiability in genomic research (Rahimzadeh et al., 2020; Selita, 2019). The perceived risks associated with AI can frighten individuals and communities from participating in genomic research, particularly if they fear that their data may be exploited or mishandled. These concerns mirror broader debates in the literature where researchers from South Africa have raised similar issues particularly the loss of control over sensitive data and the risk of unequal benefits from data use (Cengiz et al., 2024; Moodley et al., 2022; Ramsay, 2022). There is an urgent need to protect genetic information for individuals in the context of advancing AI technologies is emerging (Wang and Liang, 2024).

The sharing of genomic data presents a complex interaction between ethical and socio-cultural considerations that must be critically examined within the framework of the theory of distributive justice (Rawls, 1971). Rawls, articulated that justice involves the fair allocation of resources, rights, and opportunities in a manner that respects each person’s equal claim to basic liberties. The sharing of genomic data with commercial entities raises critical ethical questions about equity. Such arrangements may result in commercial actors profiting from genomic research without adequately compensating the individuals or communities whose data made those advances possible a concern that emerged in our own findings. These issues echo those raised in previous studies (Ashburn et al., 2000; Selita, 2019), which similarly highlight the risk that those who benefit financially from genomic research often do so without fair recognition or return to the data contributors. Fears expressed by researchers from LMICs are often dismissed by collaborators from high-income countries (Abebe et al., 2021; Atutornu et al., 2022; Serwadda et al., 2018). While genomic data is increasingly viewed as having financial value, its use by third-party companies without proper oversight or benefit-sharing undermines trust (O’Doherty et al., 2021; Roberts et al., 2017; Yotova and Knoppers, 2020). This commercialization, in the absence of equitable frameworks, exacerbates global inequalities and reinforces distributive injustice both for research participants, who see no meaningful return, and for LMIC researchers, who are often excluded from the resulting benefits and recognition. This resonates with (Kaawa-Mafigiri et al., 2023) and (Munung et al., 2017) who reported inequity and power disparities as the major drivers of unfairness and inequality in genomic collaborative research. This scenario is likened to helicopter research, or ethics dumping (Chatfield et al., 2021; Singh and Moodley, 2021), a situation where practices that would be considered unacceptable in high-income countries are carried out in LMICs in weaker or absent regulatory frameworks. Some even liken this to “ethical imperialism” or “colonialism,” particularly in collaborations where high-income countries dominate the research agenda, and researchers from LMICs often lack access to the data or its benefits. Imperialism can also be seen in situations where those advocating for unrestricted data use expect researchers from developing countries to give up all rights to their data including the right to be acknowledged for their contributions. These findings echo Moodley et al. (2022), who warned against exploitative practices in genomic research and described the phenomenon as “data imperialism.” While some participants viewed open data sharing as beneficial for scientific progress and not a form of imperialism, the consensus supports a more cautious approach, recognizing the need for fairness, and vigilance in collaborative research.

Some genetic and genomic testing companies are considering paying people to share their data with a hope that the economic incentives will lead the industry to make their genomic data available on the market (Ahmed and Shabani, 2019; Koplin et al., 2022). Generally, people have distrust for for-profit entities and are hesitant to share their data (Jackson et al., 2019; Middleton et al., 2020; Roberts et al., 2017). Our thesis guided by the theory of distributive justice is that, genomic data is a common good that is often collected and shared freely, and should therefore not be used for commercial purposes without appropriate benefit sharing with the communities from which such data was obtained.

The socio-cultural implications of genomic data sharing are profound, as cultural beliefs and values shape individual and community attitudes toward genetic information. In many cultures, discussions around genetics and health are intertwined with notions of identity, family, and community. The potential for stigma, labeling, and shame associated with genomic data sharing cannot be overlooked. Individuals may fear that their genetic information will lead to discrimination or social isolation, particularly in communities where genetic predispositions to certain conditions are viewed negatively. When comparing the discussions from dFGD2 and dFGD3 conducted at the same institution, participants in the second group expressed concerns about stigma and discrimination, whereas the third group reported no such concerns. This difference could be attributed to the demographic composition of the groups. Participants in the second group were younger individuals with aspirations for a bright future, fearing that discrimination could result in adverse consequences, such as loss of employment, if their genetic information were shared. In contrast, the third group consisted of older individuals who felt they had lived most of their lives and perceived they had less to lose. The findings agree with perspectives of Chapman et al. (2020) which indicate that individuals should not be discriminated against based on their genomic information per the Genetic Information Nondiscrimination Act (GINA; Chapman et al., 2020). There is need for community engagement in mitigating the risks of stigma and fostering a more equitable approach to collaboration and the sharing of genomic data. Engaging with communities to understand their perspectives, concerns, and cultural values is crucial for developing ethical frameworks that resonate with the individuals most impacted by genomic research. Community engagement can serve as a powerful tool for addressing distributive injustice. Involving all stakeholders in the research process (Chatfield et al., 2018), is a sign of value and form of respect, and their knowledge can be recognized and valued. Research should be with communities and not about them (Chatfield et al., 2018). The findings agree with Nankya et al. (2024) about the need for continuous community engagement.

Is it fair and morally right to sell genomic data that was freely given by participants? From the perspective of distributive justice, these practices raise serious concerns particularly when the benefits of research, including commercial gains, are not fairly shared with those who contributed their data, especially participants, and researchers from low- and middle-income countries (LMICs). In responding to the need to share data which is essential to the advancement of science, how then do we balance access to data and data sharing with meaningful participation of communities. With an increase in the sharing of genomic data, Ramsay questions whether enough is being done to ensure that there is benefit for Africa as a whole and in protecting the interests of researchers and research participants (Ramsay, 2022).

Conclusion

Ethical considerations in genomic data sharing include culturally specific continuous consent, restricted third-party data access, and addressing inequities in collaborative research are critical during the sharing of genomic data. Community engagement and stakeholder engagement is essential to navigate these challenges while ensuring transparency, accountability, and data security. The theoretical framework of distributive justice offers a valuable lens for exploring the ethical and social concerns surrounding genomic data sharing. It emphasizes fairness in the distribution of benefits and responsibilities, particularly in collaborations between high-income and LMICs. It also sheds light on how communities perceive fairness, shaped by trust and past research experiences, which are critical for promoting equitable and respectful research partnerships.

To address these issues, robust frameworks must promote equitable partnerships and benefit sharing, prevent data misuse, transparent data-sharing policies, and culturally sensitive approaches. Emphasizing privacy, informed consent, and fairness is vital. Policies should ensure the equitable distribution of research benefits while safeguarding the rights and dignity of all stakeholders.

Supplementary Material

Supplemental material for this article is available online.

Data availability statement

Details of how the analysis was done has been shared in this manuscript. The tools have been shared as supplementary items.