Decolonizing genomic data ownership: a qualitative study of legal practices and access in Uganda Authors: Deborah Ekusai Sebatta, Shenuka Singh, Jessica Oga, David Kyaddondo, John Barugahare, Sharifah Sekalala, Moses Mulumba, Erisa Mwaka Published: 2026 ABSTRACT BackgroundGenomic data sharing enhances efficiency in medical practice by facilitating faster data access and collaboration across settings. However, the same arrangements that enable efficiency could also raise legal inequalities, especially concerning data ownership and access. Yet these inequities are intensified in cross-border collaborations where power dynamics disadvantage low and middle income (LMIC) researchers.MethodsA qualitative phenomenological study was conducted in Uganda. Key informant interviews (KIIs) were conducted with 49 stakeholders working at different points across the genomic research landscape and four focus group discussions (FGDs) were conducted with people currently or previously enrolled in genomic studies, or their caretakers. The stakeholders who took part in the KIIs included genomic researchers (16), research ethics committee (REC) members (14), practicing lawyers (3), community advisory board (CAB) members (8) and national research regulators (8). A theory of decolonization was used to clarify rooted power imbalances and unfairness. Through thematic analysis six concerns were identified: (1) inadequate regulatory frameworks, (2) limited access to the shared data, (3) unclear data ownership (4) absence of benefit sharing frameworks, (5) contested intellectual property rights and (6) rethinking power asymmetries and struggles.ResultsThe wide spread perception by research participants that the Global North collaborators owned the data, eroded community trust. There was limited clarity on data ownership, with Ugandan researchers expressing differing views, on ownership, custodianship, and shared ownership leading to calls for co-ownership. In the absence of benefit-sharing frameworks to regulate the distribution of commercial value, data commercialization disadvantaged Ugandan researchers and participants, reinforcing colonial-like inequalities.ConclusionCurrent governance systems mirror colonial patterns in the control and sharing of genomic data in Uganda. These patterns are reinforced not only by cross-border power asymmetries but also by internal power dynamics within the country, including institutional hierarchies and regulatory gaps that limit local decision-making authority. Addressing this requires rethinking governance structures and promoting local or co-ownership models that reflect African interests and support decolonized approaches to data governance.Trial registrationNot applicable. This study did not involve a clinical trial.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12910-026-01517-0. Background Large-scale genomic data-sharing initiatives have become central to advancing biomedical research by generating and governing complex “big data” resources that enable cross-population analyses, accelerate discovery, and support the development of new diagnostics and treatments. In Africa, initiatives such as H3Africa, MalariaGEN, and the Roche African Genomic Program are producing and stewarding large, diverse genomic datasets drawn from multiple countries [1, 2]. Collectively, these efforts aim to improve health outcomes, strengthen local research capacity, and shift scientific leadership toward African-led research. While sharing genomic data is essential for advancing treatment and biomedical research, the lack of consistent, context-specific legal frameworks creates serious data-protection risks. The sharing of genomic data on global platforms raises a range of ethical, social, and legal concerns. Ethical concerns include the risk of re-identification, threats to privacy, and the potential for inequitable use of data. Social risks may arise in the form of discrimination, stigma, or social exclusion, particularly for individuals or communities linked to sensitive genetic information [3, 4]. Legal concerns include data security vulnerabilities, inequitable benefit-sharing arrangements, and persistent uncertainty regarding data ownership and control [5–8]. Regulatory capacity for genomic data sharing varies widely across countries. A systematic review by de Vries, Munung [9] found that only three African countries, Cameroon, Ethiopia, and Tanzania, explicitly reference data sharing in their guidance documents [9]. Moreover, the ethico-regulatory frameworks guiding genomics research in African settings are either absent, outdated, or too general to effectively guide data-sharing practices [10, 11]. For example, in Uganda, the data sharing and reuse culture has not been fully developed and policies for data sharing and reuse have not yet been fully defined especially regarding the health sector. In addition, without clear and harmonized context specific regulations, sensitive genetic information may be stored or transferred across jurisdictions with uneven privacy standards, leaving gaps in consent procedures, security safeguards, enforcement and potential risk in international collaboration [9, 11, 12]. Power imbalances have been observed not only in collaborations between researchers from the Global North and South but also in collaborations among researchers from the Global South [13–15]. The sharing of genomic data on global platforms has raised concerns among researchers from the Global South, some of whom describe these practices as a form of ethical imperialism or even “colonialism” [8, 16]. These concerns are closely connected to, though distinct from, the broader literature on digital colonialism. While digital colonialism often refers to commercially driven data extraction and platform control, genomic research is typically framed as publicly funded and ethically governed collaboration. However, similar colonial dynamics emerge when high-income partners control digital infrastructures, data repositories, analytical tools, and governance frameworks that determine how genomic data are accessed, interpreted, and reused [17].This often occurs when high-income partners and South-South collaborations set the research agenda while local scientists have little influence or access to the data they help generate. Such quiet but persistent power imbalances within global science have been described as parachute research [18]. These concerns become more troubling when studies are carried out in settings with weaker ethical oversight, allowing international collaborative research to proceed without equitable data governance arrangements among collaborators. These dynamics highlight why many feel that current data-sharing practices remain unfair and in need of reform [19]. According to Mulumba, Oga [20], the idea of decolonizing global health is grounded in efforts to address historical and ongoing power asymmetries by creating fairer and more inclusive research systems. This understanding builds on work by Abimbola and Pai, who explain that global health continues to reflect its colonial history through unequal power relations, particularly in who sets research priorities, controls resources, and whose knowledge is recognized [21]. Taken together, these perspectives describe decolonization as a call to rethink how global health research is organized so that it better reflects the voices, priorities, and expertise of those in low- and middle-income countries. Achieving this will require decolonization done in an ethical way, which involves prioritizing the perspectives of local researchers and research participants, ensuring that they benefit meaningfully from research, and challenge Western epistemic dominance [22]. It also necessitates dismantling structures that allow the Global North to retain disproportionate control over data ownership and use often at the expense of equitable participation and benefit for researchers and communities in the Global South. Central to decolonization is the meaningful involvement of those with lived experience, ensuring that researchers and communities are not only data sources but active participants in decision-making about how data are governed and used [17]. In this study it was our aim was to examine the legal asymmetries surrounding the sharing of genomic data in collaborative research, while interrogating the colonial tendencies that shape these practices. Building on previously done work [8, 23], this study will contribute to the development of a guidance paper, will push the conversation toward equity, fairness and strengthen the ethico-regulatory framework for genomic data sharing in collaborative research arrangements. Methods Theoretical perspective We draw on the theory of decolonization as a lens to challenge colonial dominance and power imbalances surrounding the sharing of genomic data in collaborative research [21, 24, 25]. Decolonization draws attention to how colonial legacies persist in contemporary research, privileging external actors’ control over knowledge, decision-making, and the distribution of benefits. We propose a decolonial agenda and examine how colonial dominance and power imbalances shape the sharing of genomic data within collaborative relationships. While the literature on digital decolonization broadly addresses efforts to undo how digital technologies, data, and infrastructures reproduce colonial power relations, our focus is on research-led genomic data sharing, where data are generated through formal ethical and collaborative processes [17, 26, 27]. This framework guided data collection, analysis, interpretation, and reporting of the findings. Applying this lens to data sharing allows us to question deep rooted power asymmetries, highlighting inequities in how data governance, ownership, and access practices can reproduce such inequities. The theory points towards approaches that promote fairness, co-ownership of data, locally grounded governance, and equitable benefit-sharing that restore agency and fairness to researchers and communities in LMICs. Regulatory context Article 79 of the Constitution of Uganda empowers the Parliament to enact primary legislation that establishes the country’s overarching legal and governance framework. These Acts of Parliament form the foundation of Uganda’s legal system, second only to the Constitution. This study distinguishes between laws, regulations, policies, and ethical guidelines, as these instruments differ in legal status and enforceability but collectively shape research governance in Uganda. Laws are enacted by Parliament, regulations are subsidiary legislation issued under an enabling Act, while policies and ethical guidelines provide normative guidance without the force of law. In Uganda, there are no standalone research regulations enacted by Parliament. Instead, research oversight is largely guided by ethical guidelines issued by the Uganda National Council for Science and Technology (UNCST) [28], while the National Drug Authority specifically regulates clinical trials. The Parliament has delegated authority through subsidiary legislation to bodies such as UNCST to operationalize primary laws. This authority is legally grounded in the UNCST Act of 1990 (Cap. 209), which mandates UNCST to develop and implement strategies for integrating research into national development, making its regulatory functions legally enforceable. However, this governance structure reveals important gaps. The enforceability of research ethics guidance rests primarily with Research Ethics Committees (RECs) and UNCST, with legal redress typically pursued through the Uganda Human Rights Commission or the High Court. Notably, Uganda lacks a specific law on genomic data protection, raising questions about whether existing institutions and ethical guidelines are sufficiently robust to prevent harmful genomic data sharing, particularly in the absence of explicit consent. Design and reporting A phenomenological approach was adopted to gain an in-depth understanding of how stakeholders experience and interpret issues related to genomic research and data sharing within Uganda’s research context. The aim of this study was to examine the legal asymmetries surrounding the sharing of genomic data in collaborative research, while interrogating the colonial tendencies that continue to shape these practices. Phenomenology was considered appropriate for this study as it enabled the exploration of participants’ perceptions, attitudes, and practices regarding genomic research and the sharing of genomic data in collaborative research settings [8, 28].The consolidated criteria for reporting qualitative research (COREQ) checklist was used to guide the transparent and comprehensive reporting of the qualitative study [29, 30]. Setting The study was conducted at research-intensive institutions in Uganda which were selected based on their sustained publication record, extensive international collaboration experience, and active engagement in genomic research [28]. The publication record was assessed through a review of publicly available peer-reviewed biomedical and genomic research publications produced by each institution over several years. These institutions host several ongoing and completed genomics studies and maintain multiple collaborations with international partners. The names of the participating research institutions will not be disclosed to protect their privacy. In addition, the study included a university categorized as an academic institution, as well as RECs responsible for approving genomic studies and providing national regulatory oversight [8]. We also selected CABs affiliated with organizations conducting genomic research, ensuring that participants had either direct or indirect engagement with genomic research activities [8]. Institutions involved in legal or policy work were identified. Sampling and population The study population comprised multiple stakeholder groups involved in Uganda’s genomic research landscape. These groups were purposively selected because they occupy distinct yet complementary roles in the generation, governance, and oversight of genomic data [8]. Stakeholders included researchers conducting genomic studies; members of RECs responsible for reviewing genomic protocols; CAB members affiliated with institutions undertaking genomic research; lawyers with experience in research-related legal matters; national research regulators; and participants in genomics research. A list of genomic researchers from Makerere University College of Health Sciences and participating institutions was compiled, and eligible researchers were invited to participate. A list of accredited RECs authorized to review genomic research was obtained from the UNCST, from which REC chairs, members, and community representatives were purposively selected [23]. Researchers identified from the list were contacted, and only four declined participation, primarily due to busy schedules, despite indicating their willingness to participate. Through snowball sampling, CAB members assisted in identifying additional participants within the group. Practicing lawyers were identified through professional networks based on their involvement in research governance. All identified stakeholders were contacted by telephone. Of the eight research institutions approached, seven agreed to participate. Five of the ten RECs with documented experience in reviewing genomic research were purposively selected based on their direct involvement in genomic study oversight, and all agreed to participate. REC members who declined participation were primarily individuals in leadership roles, citing time constraints and competing professional obligations. The study population also included individuals with prior involvement in genomic research. A sampling frame was constructed comprising adults (≥ 18 years) who had directly participated in genomic research and caretakers of individuals who had been enrolled as minors. Individuals with experience only in non-genomic research studies were excluded [8]. We purposively selected lawyers who had experience in research-related legal issues. Research team and reflexivity The research team comprised three members, led by the principal investigator (DES). All the interviews and FGDs were conducted by DES and the trained research assistants who was also the notetaker. DES is a social and behavioural scientist with training in bioethics and experience in genomic research governance. One research assistant had prior involvement in supporting the conduct of genomic research. This mix of experience provided both contextual familiarity and analytical distance. One member of the research team had prior experience working within North–South international collaborations, which provided insight into governance structures, negotiation processes, and power dynamics. While this familiarity supported deeper engagement with participants’ accounts, it also carried the risk of interpreting findings through existing assumptions about inequities in collaborative research. None of the FGD participants were known to the research team. A small number of approximately five interview participants were known in a professional capacity; however, there were no direct supervisory or hierarchical relationships. Nonetheless, shared professional networks may have influenced interactions. Given the study’s focus on power asymmetries and colonial dynamics, the team engaged in reflexive practice throughout the study. This included regular debriefing discussions after interviews and FGDs to reflect on how the researchers’ backgrounds, assumptions, and positions within genomic research might have shaped interactions with participants and the interpretation of data. Particular attention was given to potential power dynamics when engaging with different stakeholder groups, including researchers, CAB members, regulators, lawyers, REC members, and genomic research participants and caretakers. The team also critically examined how familiarity with the research context could influence probing and analysis, while the inclusion of a member without prior genomic research experience supported questioning of taken-for-granted assumptions. A standardized topic guide and neutral probing techniques were used to promote consistency across interviews. Data collection methods This study employed qualitative data collection methods, specifically KIIs, in-depth interviews (IDIs) and FGDs. The IDIs explored CAB members and REC community representatives personal experiences, perceptions and meanings in detail. The KIIs were conducted with researchers, REC members, regulators, and legal professionals who had specialised knowledge or experience relevant to the research topic, such as. The FGDs were used to capture shared experiences and collective perspectives from genomic research participants [8, 23, 28] . Study methodology An overview of the stakeholders working at different points across the genomic research landscape as well as the methods used have been provided (Table 1). Table 1Summary of the study methodologySettingReason for selectionStakeholder CategoryData collection method usedNo.University and research intense institutionsResearch intense institutionsGenomics ResearchersKIIs16Legal institutionsFocus on legal issues and researchPracticing lawyersKIIs3RECsRECs which review genomic protocolsREC members:KIIsIDIs77National research Regulatory InstitutionsBased on the regulatory oversightNational research regulatorsKIIs8CABsAffiliated to research institutions which conduct genomics researchCAB MembersIDIs8Research institutionsBased on their current and previous experienceResearch participantsFocus group discussions (FGDs)39 Table 1Summary of the study methodologySettingReason for selectionStakeholder CategoryData collection method usedNo.University and research intense institutionsResearch intense institutionsGenomics ResearchersKIIs16Legal institutionsFocus on legal issues and researchPracticing lawyersKIIs3RECsRECs which review genomic protocolsREC members:KIIsIDIs77National research Regulatory InstitutionsBased on the regulatory oversightNational research regulatorsKIIs8CABsAffiliated to research institutions which conduct genomics researchCAB MembersIDIs8Research institutionsBased on their current and previous experienceResearch participantsFocus group discussions (FGDs)39 Summary of the study methodology KIIs IDIs An overview of the methodology is presented in Table 1 Contact details of the prospective participants were obtained from the respective institutions with appropriate permission, used solely for recruitment purposes, and stored in access-restricted folders. Individuals were contacted by telephone or email, given a description of the study, and offered an interview appointment at their convenience. Data were collected between August 2023 and March 2024 [8]. A hybrid approach to data collection was used. The KIIs and IDIs were conducted either face-to-face or virtually, depending on participant availability and preference, while the FGDs were conducted face-to-face only. Semi-structured interview guides were used for the KIIs, IDIs, and FGDs. The FGD and IDI guides were translated into Luganda, the most commonly spoken language in the study area, by a professional translator with experience in translating genomic consents. To ensure linguistic and conceptual equivalence, community representatives were consulted to clarify culturally appropriate meanings of genomic-related terms, and minor revisions were made to the Luganda version accordingly [8]. The guides were piloted with participants who had ever participated in genomic studies and minor adjustments were made to improve clarity and flow. The KIIs were conducted in English, while IDIs were conducted in English and Luganda. All interviews lasted approximately 40–60 min. The use of semi structured questions encouraged in-depth discussions, allowing participants to offer nuanced insights on the legal concerns in the sharing of genomic data. A total of 49 interviews inclusive of both in-depth and KIIs were conducted. A total of four FGDs were conducted with people currently or previously enrolled in genomic studies, or their caretakers (Table 1). All FGDs and interviews were audio-recorded with the respondents’ consent to ensure accuracy in data capture. The FGDs were conducted in private spaces within the participating institutions where participants were selected from that conduct genomic research with each session lasting approximately 60–120 min. All FGDs were conducted in Luganda, the most widely spoken language in Central Uganda, to accommodate participant preferences. The KIIs, IDIs and FGDs were conducted until data saturation was achieved, defined as the point at which subsequent interviews yielded no major new codes or perspectives and merely confirmed existing themes. Saturation was assessed iteratively during data collection and analysis. The research team analysed data concurrently with data collection and held regular discussions to review emerging themes. Interview transcripts were reviewed as they were produced, and emerging codes and themes were compared across interviews. Data collection, encompassing KIIs, IDIs, and FGDs, continued until no new substantive insights emerged and subsequent data primarily reinforced previously identified patterns. Data management All audio recordings were transferred to password-protected computers and deleted from the recording devices. Recordings were de-identified prior to transcription by removing personal identifiers such as names, institutions, and specific locations. While most interviews and FGD audio recordings were transcribed verbatim in the original language by DES and two members of the research team who were involved in data collection, some files were transcribed by other trained members who were not directly involved in data collection. The transcribers worked closely with the lead researcher and were trained on study procedures, including confidentiality requirements. Where data were collected in Luganda, transcripts were translated into English after transcription [31]. To ensure accuracy and completeness, the lead researcher verified each transcript against the original audio recordings, and additional checks were conducted by members of the field team using audio recordings and field notes [32]. Analysis Data from the KIIs and FGDs were first analyzed independently using a thematic approach. This procedure involved an iterative process that involved identifying, analyzing and the interpretation of the data [33]. Figure 1 provides a schematic overview of the analytical process, illustrating the progression from data collection, transcription and familiarization through coding, code integration across stakeholder groups, and iterative theme development and refinement. Fig. 1Flow chart showing qualitative data analysis Flow chart showing qualitative data analysis Data analysis began with repeated reading of the interview and FGD transcripts to achieve familiarity with the data. Analysis was conducted using a combination of inductive and deductive thematic approaches. During the first cycle of coding, open inductive codes were generated line-by-line to capture emerging concepts without imposing predefined structure. In parallel, a deductive coding framework, informed by the study objectives and existing literature, was applied. The coding process was iterative: initial codes were compared across transcripts, refined, merged, or split as needed, and then reapplied to the data. After revising the codebook, a second cycle of coding was conducted to identify relationships among codes which were then grouped into broader categories. In practice, the theory of decolonization informed the coding process by sensitising the analysis to issues of power, control and inequity. Constant comparison across transcripts ensured consistency and analytic rigor. Finally, themes were developed by integrating inductive patterns with the deductive categories, and these themes were reviewed, refined, and validated against the raw data to ensure coherence and representativeness. A team-based approach was adopted for data analysis, involving DES and two social scientists, who developed a codebook [34]. Initial coding involved identifying recurring ideas and concepts within the transcripts. A preliminary codebook was developed and refined by three experienced researchers leading to the refinement of codes through iterative discussion rephrasing and the list of codes was cut down and merged [35]. Regular meetings were held among members of the research team to review emerging insights, and there was agreement that no new issues were arising. To enhance robustness, a small number of additional interviews and FGDs were conducted beyond this point to confirm that no new insights were emerging, after which recruitment was discontinued. The initial codebook comprised several codes derived from early transcript review. Given the diversity of stakeholder groups involved in the study, transcripts from KIIs, IDIs and FGDs were initially analysed separately to capture differences in perspective. FGDs were analysed to explore shared community-level experiences and group dynamics, while KIIs focused on individual or expert perspectives. Separate codebooks were also developed for each key informant stakeholder category, including researchers, CAB members, lawyers, regulators, and REC members. This approach enabled the research team to identify themes specific to particular stakeholder groups before integrating them into the broader analysis. Following this stage, the separate codebooks were compared and consolidated into a single master codebook. During this process, overlapping codes were merged, and related codes were grouped into broader thematic categories. The research team reviewed these themes collectively to ensure that they captured patterns across stakeholder groups while also reflecting differences in perspectives. The resulting master codebook provided the basis for identifying the final themes and subthemes reported in the study. The integration of inductive and deductive coding occurred throughout this process. The data were managed and analyzed using NVivo version 14 qualitative data analysis software. A full presentation and description of the themes is provided in the results section. Trust worthiness To enhance the trustworthiness of the analysis, multiple researchers independently coded selected transcripts and then compared coding decisions to assess consistency in code application. Discrepancies were discussed during team meetings until consensus was reached, and the codebook was further refined to ensure consistent application of codes across the dataset. No formal intercoder reliability statistics were calculated. Initial codes were generated based on recurring ideas and concepts in 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 [34]. We merged the codebooks into a single master codebook after establishing strong initial themes. The theme and subthemes were then reviewed, refined, and named in line with the research objectives. The final codebook comprised of themes which were derived from both data sources. The final master codebook consisted of themes and subthemes derived from both interviews and FGDs which reflected perspectives across the different stakeholder groups involved in genomic research governance. A summary of the coding structure is described in the Results section. Results Socio-demographics of participants The study included key informants and IDI participants drawn from a range of professional roles and institutions engaged in genomic research and governance. A total of 49 participants, from different stakeholder groups, participated in the interviews (Table 2), of which 32 were male. The majority were Ugandan, with a small number of non-Ugandan participants. The researchers included principal investigators and co-investigators with senior-level experience and qualifications at the master’s or PhD level, the REC members included chairs as well as community representatives and members. Table 2Summary of the key informants’ and IDI participants demographicsStakeholderCategoryMaleParticipantsFemaleParticipantsTotalParticipantsResearchers12416Legal representatives303REC Members10414Regulators538CAB Members368Total321749 Table 2Summary of the key informants’ and IDI participants demographicsStakeholderCategoryMaleParticipantsFemaleParticipantsTotalParticipantsResearchers12416Legal representatives303REC Members10414Regulators538CAB Members368Total321749 Summary of the key informants’ and IDI participants demographics Three FGDs were conducted with genomic research participants and one with caregivers of minors (n = 39). Both male and female participants were recruited and asked to share their perspectives on genomic data sharing; 17 participants were women (Table 3). Table 3Demographic characteristics of focus group discussion participantsFGDParticipantDescriptionAgecategoryMaleParticipantsFemaleParticipantsTotalParticipantsFGD 1Mixed group20–586511FGD 2Young adults18–275510FGD 3Parents and guardians32–606410FGD 4Mixed group20–50538Total221739 Table 3Demographic characteristics of focus group discussion participantsFGDParticipantDescriptionAgecategoryMaleParticipantsFemaleParticipantsTotalParticipantsFGD 1Mixed group20–586511FGD 2Young adults18–275510FGD 3Parents and guardians32–606410FGD 4Mixed group20–50538Total221739 Demographic characteristics of focus group discussion participants The demographic characteristics of the FGD participants are presented in Table 3. Six themes on legal concerns in the sharing of genomic data were identified guided and these included weak or inadequate regulatory frameworks, limited access to data by LMIC researchers, data ownership, absence of benefit-sharing frameworks, intellectual property challenges, and strategies addressing power asymmetries and advancing decolonization to overcome colonial tendencies. Inductive coding allowed themes to emerge directly from participants’ narratives. For example, weak or inadequate regulatory frameworks arose from accounts of gaps in national laws and challenges with enforcement. These themes reflected participants’ lived experiences of governance gaps and inequities in genomic data sharing. While strategies to overcome colonial tendencies emerged from reflections on power imbalances and calls for more equitable collaborations. Deductive coding was guided by the study objectives and informed by a decolonization lens, which was used to interrogate power asymmetries in genomic research collaborations. This approach guided the attention and focused on how participants described issues such as control over genomic data, unequal decision-making authority, and benefit sharing, power asymmetries enabling a more critical examination of how existing governance practices may reproduce structural inequities. The analysis identified several key legal concerns in the sharing of genomic data, including issues of ownership, access, and enforcement. These emergent themes are summarised in Table 4. Table 4Emergent themes on the legal concerns in genomic data sharingMain ThemeSubtheme / CategoryDescription / Illustrative PointsLimited Access to Data by Ugandan ResearchersPoor infrastructure and limited skillsetsInadequate infrastructure, training, and resources constrain local researchers’ ability to engage in data sharing.Technological disparitiesAdvanced data management and analytic tools are concentrated in high-income countries.Data ownershipUnclear ownership of data1. Ambiguity on whether data belong to research participants, researchers, institutions, or funders.Co-ownership debates2. Advocacy for shared ownership, but concerns about unequal power relations persist.Views on consent and data control3. Participants provided samples freely, researchers managed data, and donors funded the studies, complicating ownership claimsAbsence of benefit-sharing frameworksExistence of frameworksMost participants reported a lack of structured benefit-sharing mechanisms.Expected benefitsAnticipated benefits include vaccines, drugs, publications, and commercial outcomes.BeneficiariesIdentified beneficiaries include participants, research communities, institutions, and funders.Timing and discussion of benefitsStakeholders differ on when discussions should happen—before, during, or after the study.Contested Intellectual property rightsUnclear IP policiesLack of clear guidance on intellectual property rights and data use agreements create confusion.Rethinking power struggles and ensuring equityEstablishing and strengthening legal and policy frameworksCalls for clear, enforceable regulations and harmonized data-sharing policies.Co-ownership and equitable partnershipsEmphasis on balanced ownership, legal support, and frameworks that promote fairness and local benefit. Table 4Emergent themes on the legal concerns in genomic data sharingMain ThemeSubtheme / CategoryDescription / Illustrative PointsLimited Access to Data by Ugandan ResearchersPoor infrastructure and limited skillsetsInadequate infrastructure, training, and resources constrain local researchers’ ability to engage in data sharing.Technological disparitiesAdvanced data management and analytic tools are concentrated in high-income countries.Data ownershipUnclear ownership of data1. Ambiguity on whether data belong to research participants, researchers, institutions, or funders.Co-ownership debates2. Advocacy for shared ownership, but concerns about unequal power relations persist.Views on consent and data control3. Participants provided samples freely, researchers managed data, and donors funded the studies, complicating ownership claimsAbsence of benefit-sharing frameworksExistence of frameworksMost participants reported a lack of structured benefit-sharing mechanisms.Expected benefitsAnticipated benefits include vaccines, drugs, publications, and commercial outcomes.BeneficiariesIdentified beneficiaries include participants, research communities, institutions, and funders.Timing and discussion of benefitsStakeholders differ on when discussions should happen—before, during, or after the study.Contested Intellectual property rightsUnclear IP policiesLack of clear guidance on intellectual property rights and data use agreements create confusion.Rethinking power struggles and ensuring equityEstablishing and strengthening legal and policy frameworksCalls for clear, enforceable regulations and harmonized data-sharing policies.Co-ownership and equitable partnershipsEmphasis on balanced ownership, legal support, and frameworks that promote fairness and local benefit. Emergent themes on the legal concerns in genomic data sharing Theme 1: weak regulatory frameworks Several participants noted that practices such as cloud storage, biometric identification, and AI-driven analysis are evolving far faster than the laws meant to safeguard data, and many research institutions in Uganda lack legal or data-governance offices. They called for strengthening of locally led legal systems, developing of national data-governance policies and legal support in Uganda will empower local scientists and institutions to negotiate equitable data-sharing agreements, and protect their communities. Researchers, REC members, lawyers and national research regulators noted that this shift will place researchers from Uganda in the role of stewards and decision-makers, breaking from the colonial past and reclaiming authority over how genomic data is stored, accessed, and used. Moreover, there appeared to be a significant disconnect between existing policies and the actual practices of genomic data sharing, particularly in Uganda. While institutions may have general guidelines or ethical expectations in place, these often fall short of providing clear, enforceable frameworks for how data should be governed, accessed, or tracked once shared. One key issue that was highlighted was the absence of mechanisms to monitor or track data after it has been shared. Once data leaves the provider institution, there is often no way of knowing who uses it, for what purpose, or whether the terms of use are being followed. This absence of procedures and processes undermines both institutional accountability and participant protection. One REC member highlighted how this disconnect is further compounded by weak regulatory frameworks, particularly the absence of effective enforcement mechanisms. He noted that, despite contributing to data generation, local institutions often lack the regulatory authority and structures needed to ensure compliance with agreed terms governing data sharing, storage, and use, reflecting broader power imbalances in global health research.There is little systematic follow-up to monitor compliance once data have been shared. As a result, adherence to agreed data sharing terms relies largely on institutional goodwill rather than on clear oversight or enforcement structures. There are no mechanisms in place to confirm that the data is being used for what they said they would use it for (REC member 02). There is little systematic follow-up to monitor compliance once data have been shared. As a result, adherence to agreed data sharing terms relies largely on institutional goodwill rather than on clear oversight or enforcement structures. There are no mechanisms in place to confirm that the data is being used for what they said they would use it for (REC member 02). One national research regulator indicated that there is no ethico-regulatory framework to regulate genomic data-sharing, which leads to reliance on government laws. This presents a challenge, as these general legal provisions do not sufficiently address the unique ethical and governance issues associated with genomic data, resulting in regulatory gaps, weak enforcement, and potential risks to participant protection.No, in Uganda currently, we don’t have that ethical regulatory framework. What we are using really, is to read from the government laws. Then we try to see how they fit in the regulation. That is why from the beginning, I said we are lacking that. Yes, and that’s why we may need stakeholders to come in to make sure that the framework becomes very clear. It doesn’t violate the national laws and international laws, but it brings in the context based challenges we are having to see how they can be overcome and implement the framework (Regulator 5). No, in Uganda currently, we don’t have that ethical regulatory framework. What we are using really, is to read from the government laws. Then we try to see how they fit in the regulation. That is why from the beginning, I said we are lacking that. Yes, and that’s why we may need stakeholders to come in to make sure that the framework becomes very clear. It doesn’t violate the national laws and international laws, but it brings in the context based challenges we are having to see how they can be overcome and implement the framework (Regulator 5). One lawyer described the current governance landscape as fragmented, with multiple laws relating to data protection but without a comprehensive framework to coordinate their implementation. As one participant explained:What we need is a framework to organize all these pieces of laws and policies to make sense because the laws are existing. The issue is that we are not implementing them the way we are supposed to. So, the practice, and the requirements, are not just theoretical academic discussions, but they should be practical discussions that touch on people’s existence that need to be respected as much as possible. So, we lack one comprehensive framework law that addresses some of these things in detail (Lawyer 01). What we need is a framework to organize all these pieces of laws and policies to make sense because the laws are existing. The issue is that we are not implementing them the way we are supposed to. So, the practice, and the requirements, are not just theoretical academic discussions, but they should be practical discussions that touch on people’s existence that need to be respected as much as possible. So, we lack one comprehensive framework law that addresses some of these things in detail (Lawyer 01). Theme 2: limited access to data by researchers from Uganda Participants expressed mixed views on access to the shared genomic data. While data sharing has created new opportunities especially for researchers in LMICs, significant inequities in access persist. Participants noted that, historically, access to genomic data has been largely mediated by external collaborators. Although recent policy and practice shifts have begun to create opportunities for local investigators to access and use data they help generate, access remains uneven. However, many participants expressed concerns about inequitable access to shared data, which they linked to structural barriers including inadequate infrastructure, affordability constraints, limited awareness, and the dominance of externally managed digital systems. Together, these factors were seen to perpetuate power differentials and continue to disadvantage Ugandan researchers. According to participants, scenarios in which those who generate the data are excluded from decision-making about its use create mistrust and are considered unjust. National research regulators, lawyers and REC members, along with some CAB members, pointed out that power differentials place Ugandan researchers in a vulnerable position and heighten the risk of exploitation. They stressed the equity concerns in data access, affirming the rights of data generators and acknowledging their critical contextual knowledge. One respondent shared a vivid example of exclusion and externally controlled access, describing how data platforms can be unilaterally locked by distant collaborators:Just as you’re just keying in the password, the person from the other side goes and locks you out… this system now says the database has closed… Why? It is your own data (REC 12). Just as you’re just keying in the password, the person from the other side goes and locks you out… this system now says the database has closed… Why? It is your own data (REC 12). One regulator highlighted how power dynamics in collaborative research are often reinforced by funding arrangements, where sponsors from the Global North control both resources and data systems, limiting access to data for researchers in LMICs.So, as the research is being conducted, the data is uploaded to the drive that is only accessed by the sponsor and they could be having an online E-system in which data is being uploaded. So, they don’t have copies of the data. So, when the research is done, the sponsor goes away with everything, leaving no data with the researcher. And of course, this means that the researcher has no access to this data, while the sponsor has all the access to the data (Regulator 8). So, as the research is being conducted, the data is uploaded to the drive that is only accessed by the sponsor and they could be having an online E-system in which data is being uploaded. So, they don’t have copies of the data. So, when the research is done, the sponsor goes away with everything, leaving no data with the researcher. And of course, this means that the researcher has no access to this data, while the sponsor has all the access to the data (Regulator 8). These perspectives reveal how control over funding and technological infrastructure shapes power asymmetries in collaborative research, positioning participants and local researchers as data contributors while decision-making authority remains concentrated among more powerful actors. Theme 3: data ownership The data revealed ongoing controversy and uncertainty around the rightful ownership of genomic data. Perspectives on ownership varied systematically by stakeholder group, with research participants tending to view data as belonging to researchers. Some research respondents and CAB members felt strongly that genomic data belongs to the individuals from whom samples were collected. While REC members and regulators described situations in which researchers based in Global South institutions asserted ownership over genomic data, there were those who emphasized shared ownership between researchers and research participants. Some research participants who opined that the data belongs to the Global south believed that in-country researchers had the authority to speak for or make decisions on behalf of participants or communities. They viewed ownership as resting with the researcher who obtained consent and collected the samples. On the other hand, those who believed in shared ownership grounded their view in the consent process, as a contractual agreement that should clearly specify data sharing, ownership, management and access. Interestingly, some researchers did not feel a strong sense of data ownership, because they were not permitted to take samples or data with them when transferring to other institutions. They felt that the data belonged to the research institution or the funder. Moreover, while many acknowledged that, ethically, data should belong to the participant, the practical control often rests with the donor or funding agency. Researchers linked this perception to their lived experiences with funders who set conditions or imposed conditions on how data was used or shared. They felt that power imbalances, the lack of clarity, and a weak regulatory system complicate genomic data governance. Several REC members, lawyers and regulators explained that funders often reinforce their ownership of data by asserting that, without their financial backing, the research would not exist. Two participants from FGD1 expressed divergent views on data ownership and had differing understandings of their role in the research process. While some participants perceived data as belonging to researchers and funders, others asserted a sense of personal ownership grounded in their biological contribution. As illustrated below:Data belongs to the researcher and the funders because I think they need that data more than we do. I don’t think I’ll use it for anything; it is more beneficial to the researchers or the funders (P1, FGD1). Data belongs to the researcher and the funders because I think they need that data more than we do. I don’t think I’ll use it for anything; it is more beneficial to the researchers or the funders (P1, FGD1). Some genomic research participants, particularly those from FGD 2 and CAB members, expressed the view that genomic data belong to the donor, expressed the view that genomic data belong to the individual donor, as the DNA from which the data are derived originates from them.One respondent indicated that the data should be owned by the individual donor; however, in practice, it belongs to the funder, a situation that was viewed as unfair.I think that question has been coming up all the time about who owns the data and sometimes they say; the data belongs to the sponsor or to the researcher. Not even the researcher, because sometimes, we ship even the samples. I think, there’s a bit of unfairness, because data belongs to the people who have the money and those with advanced technology. Should it be like that? Participants end up participating in research and that is it. I think, there are more people who are in control of that data (CAB 08). I think that question has been coming up all the time about who owns the data and sometimes they say; the data belongs to the sponsor or to the researcher. Not even the researcher, because sometimes, we ship even the samples. I think, there’s a bit of unfairness, because data belongs to the people who have the money and those with advanced technology. Should it be like that? Participants end up participating in research and that is it. I think, there are more people who are in control of that data (CAB 08). One legal practitioner argued that the structure of many research grants reinforces unequal power relations by implicitly positioning funders as the owners of the data:These research grants we are signing which we are working under are crafted in a colonial way. They are crafted in a manner that the money giver is the owner of the data (Lawyer 01). These research grants we are signing which we are working under are crafted in a colonial way. They are crafted in a manner that the money giver is the owner of the data (Lawyer 01). Theme 4: absence of benefit sharing frameworks A key finding was the absence of a clear benefit-sharing framework, which leaves local researchers and participants susceptible to exploitation. Without formal guidance, many researchers and participants remain unaware of the potential benefits of data sharing including financial or other advantages. When they do find out, they often perceive unfairness in how benefits are distributed. This lack of a formal structure also limits the ability of local researchers to negotiate equitable collaborations. When stakeholders were asked for their views on participating in benefit-sharing discussions, many were in favour of such participation. There was a widely held belief that any benefits derived from the data should to be shared fairly with researchers and communities that have a legitimate right to be informed about prospective benefits. One researcher noted that while it seems fair for both researchers and research participants to benefit from data sharing, the prospect of monetary gain introduces an ethical dilemma, since consent forms emphasize free participation.While it feels right to share the data and for both researchers and participants to benefit. But there are technical contradictions in that because as you see, the participants have consented to voluntarily participate and now here, you’re saying we want to benefit. I think it makes the monetary benefit difficult but I think where you would make a stronger argument is equity and ease of access to the product and the benefits that arise from its use, whatever product it is, yes (Researcher 03). While it feels right to share the data and for both researchers and participants to benefit. But there are technical contradictions in that because as you see, the participants have consented to voluntarily participate and now here, you’re saying we want to benefit. I think it makes the monetary benefit difficult but I think where you would make a stronger argument is equity and ease of access to the product and the benefits that arise from its use, whatever product it is, yes (Researcher 03). Another researcher emphasized the need for discussions on benefit sharing, noting that collaborators from the Global North are open to such conversations. The participant also cautioned that these discussions should be guided by clear and simple guidelines, rather than overly restrictive regulations which might hinder the process.Bold discussions should be there, and they are healthy. So yes, we should discuss benefit-sharing, and we should have simple guidelines. We should say, ‘If any IP comes up in this work, it will be shared in a percentage of let’s say 2–5, 3–4’, we should negotiate those percentages. And at least all the funders are always willing to negotiate those percentages, because coming up with an IP is tough. So, let’s put clauses, that 1 in 1000; but if we tighten it so much for the 1 in 1000, all these 999 will not happen. Because if you make the regulations so tough, these 999 are cannot happen (Researcher 16). Bold discussions should be there, and they are healthy. So yes, we should discuss benefit-sharing, and we should have simple guidelines. We should say, ‘If any IP comes up in this work, it will be shared in a percentage of let’s say 2–5, 3–4’, we should negotiate those percentages. And at least all the funders are always willing to negotiate those percentages, because coming up with an IP is tough. So, let’s put clauses, that 1 in 1000; but if we tighten it so much for the 1 in 1000, all these 999 will not happen. Because if you make the regulations so tough, these 999 are cannot happen (Researcher 16). Participants differed on when discussions on benefit-sharing should happen. Some preferred discussing benefit-sharing during proposal development, while others thought that the discussions should be on-going throughout the study, and post-study. Some participants expressed surprise upon learning from the research team that genomic data could be commercialized, highlighting gaps in communication to the research participants. The findings showed that one of the main challenges that has put Ugandan researchers at the risk of exploitation is the fear of being judged as well as the risk of losing collaborations. Participants noted that researchers from the Global South, including many African countries, may be less inclined to voice their expectations or concerns during collaborative negotiations. Others noted that researchers often fear speaking up about benefit sharing due to cultural differences and this hesitancy was attributed to cultural norms that discourage confrontation or self-advocacy, as well as fears of being misunderstood or judged. This reluctance to speak up was viewed as a key factor contributing to exploitation of LMIC researchers in decision-making processes. When expectations are not clearly articulated, collaborators from wealthier institutions may make assumptions or proceed with arrangements that do not align with local priorities or values. A few participants expressed concern about how benefit-sharing practices should be communicated to both communities and individual participants, particularly after learning that genomic data could be commercialized. This raised questions about the ethics of framing participation as altruistic in contexts where potential downstream benefits are not clearly explained or shared. Theme 5: contested intellectual property rights (IP) There was a general feeling that IP in genomic research poses significant ethical, legal, and practical challenges in LMICs like Uganda. Researchers highlighted concerns of inequity in the absence of robust national IP policies and institutional oversight, which lays ground for exploitation. One researcher noted that the IP is usually held by the Western collaborating partners because they have the capacity to quickly analyze the sequences, quickly develop products and get the IP, capacity which LMIC researchers may not have.As you are aware, a lot of the intellectual property usually is held by the Western world because they have the capacity to quickly analyze the sequences, identify a potential drug-able mutation, develop the drug, and of course get the IP, which we don’t have. So, what would be important is beforehand, you agree with your potential collaborators that any IP that arises from this data is co-owned (Researcher 11). As you are aware, a lot of the intellectual property usually is held by the Western world because they have the capacity to quickly analyze the sequences, identify a potential drug-able mutation, develop the drug, and of course get the IP, which we don’t have. So, what would be important is beforehand, you agree with your potential collaborators that any IP that arises from this data is co-owned (Researcher 11). While some participants supported open science for the public good, others stressed the need for protective mechanisms such as the use of data agreements. Ethically, some participants rejected the idea of patenting human genomes, equating it to violation of personal identity and autonomy. These tensions reflect a broader global debate around the exploitation of community DNA and the commodification of human identity. The findings point to a pressing need for inclusive, context-sensitive IP frameworks that uphold ethical standards, promote fair collaboration, and build trust especially in vulnerable research settings. One REC member indicated that it is immoral and unfair for another person or entity to own someone else’s genomic data.While it is good to protect our local researchers, it’s just immoral, how can one’s genome be owned or patented by another party? It is, equating it to owning someone’s identity. This is viewed not just as a legal concern, but as a deeply ethical violation (REC member 04). While it is good to protect our local researchers, it’s just immoral, how can one’s genome be owned or patented by another party? It is, equating it to owning someone’s identity. This is viewed not just as a legal concern, but as a deeply ethical violation (REC member 04). One legal professional highlighted how researchers should take caution not to give away their IP rights during the signing of contracts. During the undertaking, researchers should not waive their IP rights because that is when they usually do because of the power dynamics (Lawyer 3). During the undertaking, researchers should not waive their IP rights because that is when they usually do because of the power dynamics (Lawyer 3). Theme 6: rethinking power asymmetries and struggles and ensuring equity Participants made a strong call to rethink existing power asymmetries, and struggles in genomic data sharing, and placed greater emphasis on leveled relationships with all stakeholders. With regard to data ownership, participants advocated for co-ownership models that allow all partners to have a genuine role in ownership, in how decisions about genomic data are accessed, used, and shared throughout the research process. Co-ownership was viewed as a way to address entrenched power imbalances and to ensure that local researchers and institutions are not exploited. Some participants, especially researchers, REC members and national research regulators also emphasized the need for legal support and the establishment of clear, enforceable legal and policy frameworks to govern data sharing. In addition, there were calls for explicit benefit-sharing frameworks that clearly outline how scientific, financial, and societal benefits arising from genomic research should be shared with researchers, institutions, and communities involved. Together, these measures were seen as essential for promoting fairness, accountability, and meaningful local benefit in collaborative genomic research. Such frameworks were seen as necessary to ensure that communities and researchers in LMICs are not excluded from commercialization gains or decision-making processes. The findings point to the need for deliberate efforts to decolonize data practices by ensuring that researchers in LMICs are not positioned merely as data generators, but are empowered as decision-makers in how data are accessed, used, and shared. Achieving this requires sustained investment in local research infrastructure, stronger and more responsive intellectual property frameworks, and the fostering of equitable research partnerships. Participants also pointed to emerging practices that reflect a shift towards co-ownership of data, as part of broader efforts to rethink power asymmetries in collaborative research. Two participants explained:I think you can argue that the data can be co-owned, to what extent, is where the negotiations can come in. Yes. I think it should be co-ownership because just collecting it and giving it away is not fair. I feel it’s not fair, because each of you has a big part to play, and without the other, it’s really a symbiotic relationship Without the other, you really wouldn’t do the research (REC 03). I think you can argue that the data can be co-owned, to what extent, is where the negotiations can come in. Yes. I think it should be co-ownership because just collecting it and giving it away is not fair. I feel it’s not fair, because each of you has a big part to play, and without the other, it’s really a symbiotic relationship Without the other, you really wouldn’t do the research (REC 03). Power asymmetries also need to be addressed within the relationship between researchers and research participants. In some cases, participants perceived research participation as a favour granted to them by researchers, rather than as a contribution to the research process. This perception appeared to reinforce the view that ownership of the data rests solely with researchers.Since you accepted to participate in the study and to share the data, it is for the researchers because you have accepted to help them do research. So, what you have to do is to give out the sample and the rest is theirs because they have not forced you (R1, FGD 02). Since you accepted to participate in the study and to share the data, it is for the researchers because you have accepted to help them do research. So, what you have to do is to give out the sample and the rest is theirs because they have not forced you (R1, FGD 02). One REC member also highlighted efforts to address power asymmetries through strengthened local oversight mechanisms, as one respondent explained.Data sharing should be guided and a signed data sharing agreement should be in place. The capacity of researchers should be developed to negotiate fairly. I developed a data sharing agreement for our institution. If at all you have to share the data itself, when it reaches to the destination, the principal Investigator in Uganda must know about it. We have records on, because when it reaches there, they use a particular application. I also visit the destination country to confirm that they don’t share the data with anybody outside of the original protocol (REC 14). Data sharing should be guided and a signed data sharing agreement should be in place. The capacity of researchers should be developed to negotiate fairly. I developed a data sharing agreement for our institution. If at all you have to share the data itself, when it reaches to the destination, the principal Investigator in Uganda must know about it. We have records on, because when it reaches there, they use a particular application. I also visit the destination country to confirm that they don’t share the data with anybody outside of the original protocol (REC 14). Discussion In this paper, we draw on the theory of decolonization as a lens to challenge colonial dominance and power asymmetries that shape the sharing of genomic data in collaborative research [21, 24, 36]. Ethical imperialism Our thesis is that data sharing lends itself to ethical imperialism, similar to other African scholars [8, 16, 23, 37], particularly in international collaborations where high-income countries dominate research agendas and researchers from LMICs have limited access to the data they generate or to its downstream benefits and ethical decolonization is challenging this norm. These findings challenge the dominant norm that international data and knowledge sharing is inherently ethical, showing instead that, when embedded within structurally unequal global research systems, data sharing can reproduce ethical imperialism [37]. Such imperialism is evident in expectations of unrestricted data use that require researchers from LMICs to relinquish control over their data, including rights to recognition and meaningful participation in decisions about its use, thereby reinforcing existing power asymmetries rather than promoting equitable collaboration. This, in turn, points to the need for ethical decolonization ensuring that local researchers and research participants benefit meaningfully from research while challenging Western epistemic exploitation [22]. Extending the discussion beyond ethical imperialism and the push for decolonization, our findings reveal substantial variation in how stakeholders understand and interpret the ownership of genomic data. Ownership of the shared data Ownership and genomic data governance are complicated by the involvement of multiple stakeholders, including individuals, research consortia, commercial entities, and governments, which leads to disagreements [38, 39]. There were differing views on data ownership among the different stakeholders that participated in this study, particularly research participants and community representatives. In several FGDs, many participants felt that the data belonged to the researchers, as they were the ones who had invited them to take part in the study. This belief was grounded in the trust and confidence they placed in the researchers, whom they perceived as acting in their best interests. This belief was mainly attributed to the fact that data is collected purely for scientific purposes to find solutions to diseases and conditions being studied as opposed to its eventual commercialization and its products. This trust may be misplaced because participants are often not fully informed about downstream data-sharing practices, including secondary use and commercialization, which are shaped by institutional agreements and external collaborators. Consequently, participants’ confidence in researchers may obscure underlying power asymmetries and limit awareness of how their data may be used beyond the immediate study which aligns with studies that showed that research participants were not aware that their data had already been shared and commercialized [40, 41]. From a decolonization perspective, such views reflect hierarchical research relationships in which participants are positioned as passive contributors, with limited awareness or control over how their data are subsequently used. The lack of clarity around who ultimately owns genomic data not only generated tensions among stakeholders but also exposed broader structural gaps particularly the absence of robust regulatory frameworks to guide data access, sharing, and reuse. Weak or inadequate regulatory frameworks The findings showed notable inconsistencies and gaps in the legal and ethical frameworks governing genomic data sharing, creating uncertainty about rights, oversight, and benefit sharing. Contemporary genomic data governance frameworks offer important normative and technical guidance but reveal key limitations in practice. The World Health Organization emphasizes equity, solidarity, and respect for persons [42], while GA4GH advances interoperability and responsible data sharing through technical standards [43]. The GDPR provides a robust, enforceable legal framework grounded in data protection principles, though its requirements may be difficult to operationalize in resource-constrained settings. In contrast, H3Africa offers a more contextually grounded model, emphasizing participatory governance and inclusivity [44]. However, across these frameworks, a common limitation is the lack of context-specific operationalization for LMICs and weak enforcement mechanisms. While the experiences of developing regulations for genomic data sharing in high-income countries can offer valuable guidance, such regulations should not be directly adopted without adaptation to the African context [11]. Despite revisions to the national guidelines [45], questions remain regarding the adequacy of Uganda’s framework for governing genomic data sharing. The guidelines frame data ownership not as a matter to be negotiated among collaborating partners at the outset of research. While this approach introduces flexibility and acknowledges the plurality of interests involved in data governance, it also shifts the burden of defining ownership onto actors who are often unequally positioned within research relationships. From a decolonization perspective, while this negotiated model creates space to challenge extractive ownership claims by recognising that multiple stakeholders including local researchers and institutions may hold legitimate interests in data, in contexts marked by structural inequalities, negotiation does not occur on a level playing field. More resourced actors, often from the Global North or well-funded institutions, may have greater legal, technical, and institutional capacity to shape the terms of data use and ownership, thereby reproducing existing power asymmetries under the guise of agreement. A review by Walakira et al. (2024) suggests that Uganda’s data sharing policy and regulatory framework may favour foreign partners, who often hold greater power and control over research data. These weaknesses echo broader concerns reported in the literature, where unclear policies have been linked to risks of exploitation in international collaborations [11, 46–48]. The findings align with a review by Liu et al. [10] which showed that inadequate policies and frameworks could lead to uncertainty as well as data misuse and risk of exploitation in international collaborations [10]. These weaknesses in the regulatory framework echo broader concerns reported by Cengiz, Kabanda [49], where unclear policies have been linked to risks in international research collaborations. From a decolonization perspective, ambiguity around data ownership, control, access, and IP is not merely a technical gap, but allows unequal power relations to persist and creates fertile ground for exploitation. In the absence of clear rules, decision-making often shifts toward better-resourced actors, typically from high-income settings, leaving local researchers and institutions with limited influence over data they helped generate. This uncertainty undermines trust, weakens accountability, and reinforces historical patterns of dependency. This imbalance reflects deeper structural legacies that mirror colonial patterns within scientific collaboration. Power asymmetries Notably, several participants described instances where researchers from Global South institutions asserted ownership over data and even over participant communities and positioned themselves as having authority to speak for or make decisions on behalf of participant communities [50, 51]. This prompts an important question, are power dynamics typically attributed to North–South collaborations or are they also present within South–South contexts? Importantly, are the voices of the research participants passive and have they been silenced? In this study, such imbalances appeared to limit participants’ ability to negotiate or assert claims over their data, with some expressing a sense of dependence on researchers. This dynamic may have contributed to the perception that data ownership resides solely with researchers from the Global North. From a decolonization perspective [22, 52], these findings highlight the need to critically interrogate hierarchical research relationships that position participants as passive contributors rather than active stakeholders with legitimate claims in data governance. Our findings suggest they may be and it is not surprising that researchers may dominate over participants; framing data as their property rather than as a shared resource that originates from the community. The findings align with the views of Sorbie et al. who asserts that when researchers dominate interests in control of data, it reveals how ownership claims are embedded within hierarchical knowledge systems that may sit uneasily with open access ideals and more equitable models of data sharing [53]. When researchers in Global South institutions claim ownership over data, it adds another layer to the already uneven power dynamics of genomic data sharing and can leave the contributing communities feeling sidelined. This pattern can unintentionally reproduce colonial-like structures, where data are taken and used to benefit researchers or institutions rather than the communities from which the data originate. These imbalances may constrain participants’ agency, shaping perceptions that data ownership rests exclusively with researchers. Such dynamics reflect entrenched hierarchical logics that decolonial approaches seek to challenge, underscoring the importance of recognizing participants as active contributors with legitimate claims over the use and governance of their data. These forms of authority can enable researchers to act as gatekeepers over communities and knowledge production. As a result, practices associated with colonial research such as extraction of data and speaking on behalf of communities may be reproduced locally, even in the absence of direct involvement from Northern actors [54, 55]. The findings on institutions being custodians of the research data align with the Uganda Data Protection Act [56] and the Chinese model which advocates for custodianship and not ownership of genomic data and this is based on cultural values and fostering trust. Yet, these suggestions are rarely implemented in practice [57]. A study conducted in Africa states that data ownership is often unclear with many researchers expressing uncertainty about who ultimately owns research data once it is generated and shared [58]. Persistent ethical debates have consistently advocated Global South researchers to assert custodianship and, in certain circumstances, to grant the state ownership of data [58–61]. Without explicit benefit-sharing frameworks, these asymmetries risked reinforcing extractive practices, where gains flow upwards to better-resourced partners while local communities receive minimal return. Absence of benefit sharing frameworks Moreover, participants repeatedly emphasized the importance of equity and the fair distribution of benefits arising from shared genomic data. However, they questioned how such equity could be achieved in the absence of a clear benefit-sharing framework, especially in collaborations often shaped by power imbalances. The result is a power imbalance that mirrors colonial patterns, where control over data and its benefits stays outside the communities that generate it [62]. These concerns directly align with decolonization critiques, as the lack of benefit-sharing frameworks and unequal control over data reproduce extractive practices that privilege external institutions and marginalise local communities mirroring colonial patterns that decolonization efforts seek to redress. Participants noted that researchers from the Global South may be less inclined to voice their expectations or concerns during collaborative negotiations. In particular, some observed that researchers often hesitate to speak openly about benefit sharing due to cultural norms that discourage confrontation or self-advocacy, as well as fears of being misunderstood or judged. In the Ugandan context, this reluctance is compounded by the absence of a clear benefit-sharing framework, particularly in collaborations where high-income country partners dominate the research agenda. Without formal guidance, many remain unaware of the benefits, some of which are profits or advantages gained from shared data, and when they do find out, they often perceive unfairness in how benefits are distributed [63]. Participants recognised that when neither benefits nor access are equitably negotiated, the result is a system in which control and the advantages it brings remains concentrated with dominant partners. As a result, researchers from LMICs may have limited access to the data and few opportunities to share in the benefits arising from downstream use, reinforcing existing power imbalances. Such scenarios could be liked to colonialism in collaborative research which aligns with research conducted by Ekusai et al.; Moodley et al.; Sekalala et al. [8, 16, 17]. This thesis therefore argues for a fair and context-sensitive approach to benefit sharing that extends beyond financial or commercial gains to include meaningful access to data, capacity strengthening, authorship, and shared decision-making and co-ownership of data. The findings on the absence of a benefit sharing framework points to research conducted in Africa [5, 64, 65] which showed that North–South collaborations often distribute benefits unevenly where institutions in high-income countries typically secure the greatest gains, while local researchers and the very communities that contribute the data are left with little recognition or reward and may even face stigma. Limited access to the shared data Consequently, there is lack of evidence on how researchers safeguard respondent interests and negotiate equitable access and reuse. For decades, advanced sequencing platforms, high-end computing, and specialized training were concentrated in wealthier countries while collaborative projects extracted data from LMICs without equivalent investment in local capacity [66]. The result is a structural dependency that leaves local researchers negotiating from a weaker position and vulnerable to inequitable agreements and rethink the power asymmetries. Rethinking power asymmetries and struggles and ensuring equity Recognizing the need to rethink the historical context shifts the narrative from blaming individuals or institutions for lacking skills to understanding how global research practices and the unequal distribution of resources have produced and maintained these disparities. Addressing power asymmetries and the contestations surrounding data ownership necessitates a shift towards co-ownership models that redistribute authority and recognize multiple claims to data which resonates with McGuire et al. [67] Our findings align with Mulumba, Oga [20] who emphasize the importance of implementing strategies to address power imbalances and dismantle neocolonial dynamics in collaborative research. They also resonate with Munung, de Vries [68], who advocate for justice, fairness, and equity, particularly in contexts where LMICs contribute data but are often excluded from meaningful decision-making. Without meaningful improvements in the capacity of genomic research stakeholders in the Global South, alongside supportive policies and effective monitoring mechanisms, genuine co-ownership risks remaining more aspirational than achievable. Strengthening governance systems that promote equitable access and oversight is therefore essential, including the use of transparent data access processes, locally anchored oversight structures, and benefit-sharing frameworks that align with decolonization principles. Addressing this problem requires more than short-term training; it demands deliberate investment in negotiation skills, policy reform, and governance models that shift power and decision-making to local actors. This study contributes to the literature on genomic data governance and decolonization in three important ways. First, it provides novel empirical evidence from a LMIC context, illustrating how governance gaps are experienced in practice across multiple stakeholders. Second, it advances decolonial scholarship by showing how structural inequities are reproduced not only through research collaborations but also through the legal and regulatory mechanisms that govern data sharing. Third, it offers actionable policy insights by identifying concrete institutional and regulatory weaknesses including gaps in enforcement, access, and benefit sharing that must be addressed to achieve more equitable genomic data governance. Strengths and weaknesses The main strength of this study was the involvement of various categories stakeholders in genomic research in Uganda. The theory of decolonization provides an in-depth understanding of the legal issues with a focus on clarifying rooted power imbalances and unfairness in collaborative research. While the findings offer in-depth insights into the study context, their generalizability to other settings are limited, underscoring the need for further research across diverse contexts. Conclusion Current governance systems across many LMIC settings continue to reflect colonial patterns in the control and sharing of genomic data, with decision-making authority and downstream benefits often concentrated outside local institutions. These patterns are reinforced not only by cross-border power asymmetries but also by internal power dynamics within the country, including institutional hierarchies and regulatory gaps that limit local decision-making authority. Recommendations Policy level classifications We advocate the development of benefit-sharing frameworks in Uganda that guarantee that the communities and researchers from LMICs are protected and not excluded from decision-making or the gains from commercialization of the data. Ensuring transparency in negotiations and clarity around IP can help prevent exploitation. There is a need to put in place practical protections that combine clear contractual terms specifying which country’s law applies and where disputes will be resolved. While such contractual provisions determine the applicable legal framework, their effectiveness ultimately depends on enforcement capacity. Given that power imbalances often shape negotiations, it is therefore essential to strengthen ethico-regulatory frameworks for data sharing, reinforce national IP and data protection laws, and establish mandatory benefit-sharing requirements to reduce the risk of exploitation. In addition, robust monitoring mechanisms are needed to track how shared genomic data are accessed, reused, and commercialized over time, ensuring accountability and compliance with agreed terms. Such reforms should be grounded in local contexts, reflect African values, and promote decolonized approaches to data governance that enhance fairness, accountability, and public trust across collaborative genomic research. Institutional level classifications There is limited evidence on how researchers safeguard the interests of research participants, or how they negotiate equitable access to data and data reuse within frameworks that may be insufficient and within organisational cultures that often discourage confrontation. Improve internal policies and practices around ethical data sharing and equitable collaboration and ensure LMIC researchers are treated as equal partners across all stages of research (conceptualisation, design, implementation, and data governance). Collaborative nature recommendations As part of efforts to decolonize data sharing, we recommend promoting models of co-ownership and strengthening local data infrastructure. Decolonizing practice requires deliberate efforts to ensure that LMIC researchers are not treated merely as data generators in externally designed projects, but recognized as equal partners in the conceptualization, planning, and implementation of research, as well as in decisions regarding data access, use, and sharing. Emphasizing co-creation of knowledge is critical to addressing power imbalances and ensuring that collaborative research reflects local priorities and expertise. This includes investing in local infrastructure, reinforcing IP frameworks, and cultivating equitable research partnerships. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.