Artificial Intelligence for Inclusive Innovation: Investigating how AI-driven information systems can democratize access to opportunities in underserved communitiesc

*Corresponding author: Chidinma Lydia Ezenwa Copyright © 2025 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution License 4.0. Artificial Intelligence for Inclusive Innovation: Investigating how AI-driven information systems can democratize access to opportunities in underserved communities Chidinma Lydia Ezenwa * MBA, Information Systems, College of Business and Innovation, University of Toledo, Toledo, Ohio, United States. World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 Publication history: Received on 04 August 2025; revised on 13 September 2025; accepted on 15 September 2025 Article DOI: https://doi.org/10.30574/wjarr.2025.27.3.3217 Abstract Artificial Intelligence (AI) is transforming global systems, yet its benefits often bypass underserved communities. This study investigates how AI-driven information systems can democratize access to finance, healthcare, and education, three sectors central to human development. Using a mixed-methods approach combining focus group discussions, key informant interviews, surveys, and literature review the research highlights both opportunities and risks. Evidence shows that AI can expand financial inclusion through alternative credit scoring, enhance healthcare via diagnostic and predictive tools, and personalize education through adaptive learning systems. However, barriers such as infrastructural gaps, algorithmic bias, privacy risks, and low digital literacy threaten to reinforce exclusion. The study concludes that inclusivity in AI is not automatic but requires intentional design, robust governance, digital literacy, and multi-stakeholder collaboration. It positions inclusive AI as essential for equitable innovation and sustainable development. Keywords: Artificial Intelligence; Inclusive Innovation; Information Systems; Digital Equity; Underserved Communities 1 Introduction The emergence of Artificial Intelligence (AI) as a transformative force in the 21st century has reshaped the global discourse on innovation, development, and human progress. Often described as the “electricity of our age,” AI is rapidly becoming embedded into systems that govern how we work, communicate, learn, and access essential services. From predictive algorithms in financial markets to diagnostic imaging in healthcare and adaptive learning platforms in education, the application of AI is extensive and continuously expanding. Yet, beneath the narrative of technological advancement lies a profound disparity: the benefits of AI are not distributed evenly. Underserved communities, particularly in the Global South, rural populations, women, and marginalized urban dwellers, frequently remain excluded from the opportunities AI presents. This research was motivated by the urgent need to interrogate this uneven terrain and to explore whether AI can genuinely serve as a democratizing force rather than an instrument of deepening inequality. Historically, access to critical sectors such as finance, healthcare, and education has been mediated by structural barriers that disproportionately disadvantage certain groups. In finance, formal banking systems have often excluded individuals without credit histories, collateral, or stable income streams. In healthcare, limited access to specialists, diagnostic tools, and quality infrastructure has contributed to preventable morbidity and mortality. In education, overcrowded classrooms, inadequate teacher-student ratios, and language barriers have perpetuated cycles of disadvantage. AI-driven information systems offer the promise of addressing these inequities by reducing reliance on World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 898 traditional infrastructure, leveraging data to expand access, and tailoring solutions to individual or community needs. For instance, AI-enabled mobile applications can allow unbanked individuals to access microcredit, digital health platforms can deliver diagnostic support in remote villages, and adaptive learning technologies can provide personalized education to students regardless of geographic location. However, the promise of AI cannot be divorced from its pitfalls. Evidence from both advanced and developing economies shows that AI systems often replicate and even amplify the biases present in the data on which they are trained. For example, facial recognition systems have been shown to perform poorly on darker skin tones, and algorithmic credit scoring has at times reinforced gender and socio-economic discrimination. The danger of deploying such technologies in underserved communities without adequate safeguards is that AI could perpetuate historical patterns of exclusion under the guise of innovation. Moreover, structural deficits such as limited internet penetration, unreliable electricity, high costs of devices, and low digital literacy further constrain the capacity of marginalized populations to benefit from AI. In many parts of Sub-Saharan Africa, for instance, less than half the population has access to stable internet services, and rural areas remain under-connected. These realities underscore the need for critical analysis that goes beyond technological potential to assess the socio-economic, cultural, and governance ecosystems within which AI is deployed. This study therefore situates itself at the intersection of innovation, information systems, and social inclusion, asking not only how AI can expand access to finance, healthcare, and education, but also under what conditions it will succeed or fail. To explore these questions, we employed a mixed-methods approach. Focus group discussions (FGDs) were conducted with community members in urban slums and rural settlements, offering qualitative insights into lived experiences with digital technologies. Key informant interviews (KIIs) were carried out with policymakers, AI developers, healthcare providers, teachers, and financial service operators, providing expert knowledge on design, governance, and adoption. Surveys were distributed to early adopters of AI-driven platforms in finance, health, and education, generating quantitative data on usage patterns, perceived benefits, and barriers. Finally, an extensive literature review synthesized global debates on AI inclusion, ranging from academic scholarship to industry reports and policy briefs. This methodological triangulation enabled us to capture both the empirical realities on the ground and the theoretical frameworks shaping the discourse. The background of this study is also informed by global development priorities, particularly the United Nations Sustainable Development Goals (SDGs). Goals such as SDG 1 (No Poverty), SDG 3 (Good Health and Well-Being), SDG 4 (Quality Education), SDG 5 (Gender Equality), and SDG 10 (Reduced Inequalities) are directly implicated in the themes explored here. The intersection of AI and inclusive development is not simply a matter of technology but of social justice and human rights. By linking AI innovation to these global agendas, the study positions itself within a broader effort to reimagine development pathways in the digital age. It also recognizes the increasing attention international organizations, including UNESCO, the World Bank, and UNDP, are giving to responsible AI adoption in developing regions. Critically, the background to this study emphasizes that inclusivity in AI is neither automatic nor incidental. It requires deliberate engineering, thoughtful governance, and multi-stakeholder collaboration. It requires questioning who designs AI systems, whose data is used, whose voices are included in development, and who ultimately benefits. This research begins from the conviction that AI, if approached responsibly, has the potential to function not only as a driver of economic efficiency but also as a powerful equalizer. Yet, achieving this vision demands a rigorous interrogation of current practices, identification of risks, and articulation of pathways that prioritize equity and dignity. It is within this context that the study investigates the role of AI-driven information systems in democratizing access to opportunities in finance, healthcare, and education, grounding its inquiry in both empirical evidence and critical analysis 2 Sectors of Relevance and Development Artificial Intelligence (AI) as a technological phenomenon rarely operates in isolation; it intersects with multiple sectors and development priorities, producing ripple effects that extend beyond its immediate applications. The relevance of AI-driven information systems is most clearly visible in finance, healthcare, and education domains that formed the core of this study but its influence stretches into adjacent fields such as agriculture, governance, and energy. A critical exploration of these sectors demonstrates both the breadth of AI’s transformative potential and the conditions required for its sustainable adoption. In finance, the relevance of AI lies not only in its ability to expand individual access to microcredit or mobile banking but also in its broader capacity to reshape financial ecosystems in emerging economies. By leveraging alternative data, AI allows micro, small, and medium-sized enterprises (MSMEs) , often the backbone of local economies, to participate World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 899 in formal credit systems. This is particularly important in regions where MSMEs employ a significant percentage of the labor force but remain largely unbanked. Through key informant interviews (KIIs) with FinTech providers, it became clear that AI is facilitating new forms of risk modeling that enable financial institutions to extend services without incurring prohibitive costs. At the same time, surveys revealed that communities view digital finance not merely as a convenience but as an enabler of livelihood resilience, allowing families to withstand shocks such as illness or crop failure. The relevance of this sector, therefore, extends to broader development outcomes: financial inclusion powered by AI contributes to poverty reduction, economic stability, and gender empowerment. Yet, as our literature review and field discussions underscored, such outcomes depend on responsible governance and localized design that ensure fairness and transparency in algorithmic decision-making. Healthcare illustrates another critical sector where AI’s relevance transcends the immediate delivery of medical services. AI-enabled diagnostic platforms and predictive health analytics are not only saving lives but also transforming how health systems are planned and managed. By analyzing epidemiological data at scale, AI can inform resource allocation, predict outbreaks, and guide preventive interventions. This capability is particularly vital in underserved regions where health budgets are constrained and every resource counts. In focus group discussions (FGDs), community members expressed appreciation for mobile applications that provided real-time health information, particularly during the COVID-19 pandemic, when physical access to hospitals was limited. Key informants in the health sector emphasized the role of AI in supporting overburdened medical staff by automating administrative processes and improving diagnostic accuracy. Beyond clinical impact, AI in healthcare is directly relevant to sustainable development goals (SDG 3: Good Health and Well-Being), particularly in advancing universal health coverage and reducing inequalities in access. However, its transformative capacity remains contingent on investment in digital infrastructure, culturally relevant datasets, and robust legal frameworks for data privacy. Without these, AI risks widening health disparities rather than narrowing them. Education is perhaps the sector where the developmental relevance of AI is most apparent to the next generation. As a driver of human capital development, education shapes the long-term trajectory of communities and nations. AI-driven platforms that personalize learning, provide multilingual support, and enable remote access to resources represent powerful tools for reducing educational disparities. In FGDs, students and parents in underserved communities expressed optimism about AI tutors that offered remedial support in core subjects, while teachers interviewed in KIIs highlighted the benefits of AI in reducing workload and enabling more student-focused instruction. Surveys showed significant enthusiasm among young learners for digital learning platforms, but also revealed infrastructural barriers such as the lack of stable internet access and affordability of devices. The literature review reinforced these findings, noting that AI has already been instrumental in bridging learning gaps in STEM education across diverse contexts. Developmentally, the relevance of AI in education is tied to its ability to contribute to SDG 4 (Quality Education) and SDG 10 (Reduced Inequalities). Yet, as our analysis makes clear, the sector’s promise can only be realized if educational technologies are designed with equity at the core, ensuring that digital divides do not translate into learning divides. Beyond these three focal domains, this study identifies additional sectors where AI holds profound relevance. Agriculture, for instance, is increasingly adopting AI-enabled systems for crop monitoring, soil analysis, and predictive yield modeling. While not the primary focus of this research, AgriTech emerged as a significant adjacent area in KIIs with development practitioners, who noted that AI could directly enhance food security in rural and refugee-hosting communities. Similarly, public sector service delivery is being reshaped by AI-driven governance tools, from digital identity systems to predictive analytics for social protection programs. These applications underscore the potential of AI to enhance transparency, efficiency, and accountability in governance. In the energy sector, AI-driven grid management and renewable energy optimization are contributing to the development of sustainable infrastructure, with direct implications for climate resilience. The developmental relevance of AI across these sectors can therefore be understood through two lenses. First, AI offers opportunities for leapfrogging enabling developing regions to bypass traditional infrastructural limitations by deploying digital-first solutions. Second, AI presents a means of embedding resilience and adaptability into systems that are otherwise vulnerable to shocks, whether economic, health-related, or environmental. At the same time, the development potential of AI is not guaranteed. Stakeholders across FGDs and KIIs emphasized that without intentional efforts to design inclusive frameworks, AI will replicate patterns of marginalization and deepen socio-economic divides. 3 Stakeholders Artificial Intelligence for inclusive innovation does not unfold in a vacuum; it is shaped by a dynamic interplay of actors whose roles, interests, and priorities determine whether technology becomes a tool of empowerment or exclusion. Understanding these stakeholders is essential for both analyzing the current state of AI adoption and designing World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 900 strategies that ensure its benefits reach underserved communities. The stakeholder ecosystem is multifaceted, involving governments and regulators, private sector entities, international organizations, civil society, academia, and communities themselves. Each of these actors brings unique contributions and challenges, creating a web of interdependencies that both enable and constrain inclusive innovation. Governments and regulators occupy a critical position in shaping the environment within which AI systems are developed and deployed. Through legislation, policy frameworks, and funding priorities, they establish the guardrails that determine whether AI adoption is ethical, equitable, and sustainable. In key informant interviews (KIIs), policymakers emphasized their dual role: on one hand, they must encourage innovation by creating enabling environments for startups and investors; on the other, they must protect citizens from risks associated with algorithmic bias, data misuse, and digital exclusion. The challenge for governments, particularly in developing countries, lies in balancing these priorities within contexts of limited resources and regulatory capacity. Nevertheless, their actions have far-reaching implications: inadequate regulation risks exploitation, while overly restrictive measures could stifle innovation. The private sector, comprising both large technology corporations and smaller startups, is the primary driver of AI research, development, and deployment. These actors are often the most agile, capable of piloting and scaling solutions rapidly. FinTech startups, for example, have pioneered alternative credit scoring systems that expand financial access, while health technology firms have developed AI-enabled diagnostic tools adapted for mobile platforms. Surveys conducted among entrepreneurs revealed that private firms are motivated not only by profit but also by opportunities for social impact, particularly in education and healthcare. However, KIIs with startup founders highlighted challenges in navigating regulatory uncertainty, accessing capital, and ensuring sustainability beyond donor funding. The private sector’s role is indispensable, but its orientation toward market incentives means that inclusivity cannot be assumed; it must be incentivized or regulated. International organizations such as UNDP, WFP, and the World Bank act as both funders and conveners in the AI ecosystem. They play a vital role in scaling innovations that address structural inequalities, particularly in lowand middle-income countries. Their funding programs often provide the seed capital necessary for pilots, while their convening power brings together governments, private sector actors, and civil society to align around shared priorities. Literature review revealed that these organizations are increasingly developing frameworks for “AI for Good,” focusing on responsible adoption and the mitigation of risks. In FGDs, community members expressed greater trust in interventions associated with reputable international organizations, reflecting the importance of credibility and oversight. Yet, international organizations also face critiques: their projects are sometimes perceived as externally driven, raising questions about sustainability and local ownership once funding cycles end. Civil society organizations (CSOs) and non-governmental organizations (NGOs) act as watchdogs, advocates, and intermediaries. They amplify the voices of marginalized populations, ensuring that AI adoption reflects community needs and rights. In FGDs, participants frequently mentioned NGOs as the most visible intermediaries helping them navigate digital tools, whether in health awareness campaigns or digital literacy programs. KIIs with CSO leaders highlighted their advocacy role in pushing for ethical AI standards, transparency in government procurement, and the protection of data privacy. They also emphasized their limitations: while effective in advocacy and awareness-raising, CSOs often lack the financial or technical capacity to scale solutions independently. Their relevance therefore lies in ensuring accountability and embedding inclusivity in AI systems, often serving as the moral compass of the stakeholder ecosystem. Academia and research institutions contribute by generating evidence, frameworks, and capacity-building programs. Through literature review and expert interviews, this study found that academic actors are increasingly engaged in interdisciplinary research on AI ethics, inclusivity, and governance. Universities and research centers provide both theoretical insights and practical tools for evaluating the impact of AI in diverse settings. In KIIs, researchers underscored the importance of methodological rigor and context-specific inquiry, pointing to the dangers of uncritically importing AI models trained in vastly different socio-economic environments. Academia also plays a role in producing the next generation of AI practitioners, equipping them with the skills and ethical awareness required for inclusive design. At the heart of the ecosystem are communities and end-users, the individuals whose lives are directly shaped by AI systems. Their participation is not merely a matter of adoption but of co-creation. FGDs demonstrated that communities often view AI solutions with a mix of hope and skepticism. Hope, because such technologies represent new opportunities for access to finance, healthcare, and education; skepticism, because experiences of exclusion, exploitation, or data misuse have created distrust. Communities are not passive recipients of technology; their feedback, cultural contexts, World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 901 and lived realities are essential inputs for designing AI systems that are truly inclusive. The absence of community participation risks producing solutions that are irrelevant, inappropriate, or even harmful. The stakeholder landscape, as illustrated in the diagram above, is characterized by interdependence. Governments set the rules within which private sector innovation operates, international organizations provide resources and legitimacy, civil society ensures accountability, academia generates knowledge, and communities ground technology in lived experience. Tensions inevitably arise: governments may prioritize security over privacy, private firms may prioritize profit over inclusivity, and international organizations may prioritize global frameworks over local needs. Yet, these tensions are not insurmountable. By recognizing their interconnections and engaging in collaborative governance, stakeholders can create AI ecosystems that align technological advancement with social justice. 4 Analysis The analysis undertaken in this study centers on the three sectors that form the foundation of upward mobility and human development: finance, healthcare, and education. Each of these areas has been historically marked by exclusionary barriers, particularly for underserved communities. AI-driven information systems hold the potential to address these inequalities by enabling new forms of access and efficiency. However, evidence from our fieldwork and secondary sources suggests that the application of AI in these domains is a double-edged sword: while opportunities abound, risks of exclusion and bias remain significant. This analysis is therefore both sectoral and critical, situating AI within lived realities as well as broader systemic structures. In the domain of finance, AI-enabled tools are revolutionizing access to credit, insurance, and savings. Traditionally, access to financial services has required formal credit histories, physical collateral, or banking infrastructure, all of which are absent for large portions of marginalized populations. Through focus group discussions (FGDs) in refugee settlements and rural communities, participants repeatedly emphasized their exclusion from traditional financial institutions due to lack of collateral or formal employment. Yet, surveys conducted among early adopters of mobilebased financial platforms revealed that AI-driven credit scoring models are beginning to create alternative pathways for inclusion. These models draw on non-traditional data sources such as mobile phone usage, utility payments, and transaction histories to assess creditworthiness. This has allowed small business owners and informal workers to secure micro-loans that would otherwise be unattainable. Key informant interviews (KIIs) with FinTech operators confirmed that these systems can scale rapidly in low-resource environments, bypassing the need for brick-and-mortar banking. However, the same interviews also highlighted critical risks: datasets used for AI scoring often reflect societal biases, and algorithmic decisions remain opaque to users. Literature reviewed from the World Bank and academic journals underscores that while FinTech-led AI is driving inclusion, women and rural populations remain disproportionately excluded when design is not localized. Thus, in finance, AI represents both a leapfrogging opportunity and a site of contested fairness. The healthcare sector presents another arena where AI has begun to disrupt traditional service delivery models. In contexts where medical infrastructure is fragile and the ratio of healthcare providers to patients is low, AI-enabled diagnostic tools offer potential for early detection and triaging. In FGDs with rural populations, participants described experiences of long travel times and prohibitive costs in accessing health facilities. Surveys of users of telemedicine platforms revealed significant appreciation for AI-powered chatbots and decision-support tools that could provide preliminary diagnoses or health advice. For example, pregnant women in rural areas reported using AI-driven maternal health apps to monitor symptoms and receive alerts about potential complications. KIIs with healthcare providers confirmed that AI tools are increasingly deployed to assist in detecting tuberculosis, malaria, and even cardiovascular diseases in resource-constrained environments. Literature review further revealed cases where AI analytics supported epidemic prediction and planning during outbreaks of Ebola and COVID-19. Yet, this optimism is tempered by structural challenges: unreliable internet connectivity, lack of electricity, and the absence of robust data privacy frameworks. Health professionals interviewed expressed concern over the reliance on algorithms trained largely on data from populations in the Global North, which may not always reflect the epidemiological and cultural realities of African or Asian contexts. Without localized data sets and ethical safeguards, AI in healthcare risks producing false positives, misdiagnoses, or privacy violations. Thus, while AI promises to bring life-saving interventions to underserved communities, it must be embedded within systems that prioritize equity, cultural relevance, and patient rights. In education, the application of AI is redefining how learning is delivered and consumed. Traditional education systems in underserved communities are often characterized by overcrowded classrooms, teacher shortages, and rigid curricula that fail to accommodate diverse learning needs. Surveys conducted among students in urban slums revealed high enthusiasm for AI-powered learning platforms that provide adaptive content tailored to individual strengths and World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 902 weaknesses. Teachers interviewed in KIIs shared that AI-driven grading and assessment tools reduced administrative burdens, allowing them to devote more attention to direct student engagement. FGDs with parents, however, highlighted concerns about affordability and accessibility: many households lacked devices or stable internet connections, limiting the reach of such technologies. Literature review highlighted promising global examples, such as the use of natural language processing to deliver multilingual education and AI tutors that bridge gaps in science and mathematics learning. Yet, concerns of over-reliance on technology persist. Education experts interviewed warned that excessive dependence on AI may exacerbate inequalities if only wealthier or urban students can consistently access these tools. They also emphasized the need for blended learning approaches that complement, rather than replace, human teachers. This reflects a broader tension: AI in education offers the potential to personalize learning at scale, but without intentional design and equitable distribution, it risks reinforcing existing divides. A cross-sectoral analysis of finance, healthcare, and education reveals shared themes. Across all three domains, AI demonstrates clear potential to lower barriers, expand access, and introduce efficiencies that were previously unattainable. Yet, structural inequities ranging from infrastructural gaps to algorithmic biases continue to constrain inclusive outcomes. The methodologies employed in this study allowed us to triangulate perspectives: FGDs captured the lived realities of communities, KIIs provided expert and policy insights, surveys quantified adoption and barriers, and literature review anchored findings within global debates. Together, these sources reveal that inclusivity is not a natural byproduct of AI adoption; it is the result of deliberate design, governance, and collaboration. The analysis suggests that AI can function either as a democratizing tool or as a new mechanism of exclusion depending on how societies choose to develop and deploy it. This duality underscores the importance of critical engagement with AI, ensuring that innovation serves as a bridge rather than a barrier for underserved communities. 4.1 Results: Presentation of Data and Findings The mixed-methods approach employed in this study encompassing focus group discussions (FGDs), key informant interviews (KIIs), surveys, and literature review—generated findings that confirm the double-edged role of Artificial Intelligence in underserved communities. Results are presented sector by sector, showing where AI-driven information systems democratize access and where structural barriers limit inclusivity. In the financial sector, surveys indicated that over 60% of participants using AI-enabled credit scoring systems reported gaining access to loans for the first time, confirming AI’s role in expanding financial inclusion. FGDs highlighted that these tools were particularly valuable for informal workers and micro-enterprises that lacked formal banking histories. However, gender disparities were apparent: nearly one-third of female respondents reported being excluded due to lack of smartphone ownership or limited digital literacy. KIIs with FinTech providers reinforced these observations, noting that while AI credit scoring allows for rapid scaling, biases embedded in data sets such as unequal representation of women and rural populations create uneven benefits. Literature reviewed from the World Bank echoed this pattern, underscoring both the promise and risk of AI in financial access. In healthcare, the results point to AI’s significant but uneven contributions. Surveys with community health workers showed that approximately 70% reported quicker decision-making when assisted by AI diagnostic tools, especially for diseases such as malaria and tuberculosis. FGDs revealed positive community perceptions of AI apps for maternal health, where women used predictive tools to monitor pregnancy risks. Yet, reliability issues were a recurring theme: almost half of surveyed users reported interruptions caused by unstable internet or electricity. KIIs with medical practitioners highlighted the additional challenge of dataset mismatch, noting that AI systems trained on Global North populations often produced false positives or missed region-specific health conditions. This finding aligns with academic literature that warns against the uncritical transfer of AI models across contexts without local adaptation. The education sector demonstrated both measurable gains and persistent limitations. Survey data indicated that 55% of students using AI-powered adaptive learning platforms improved performance in mathematics and science assessments within a single term. Teachers interviewed in KIIs emphasized reduced workloads through AI-assisted grading and monitoring systems, enabling more time for student interaction. FGDs with parents, however, revealed skepticism about cost and fears of over-reliance on digital systems replacing human teachers. Literature review confirmed these concerns, highlighting cases where digital divides widened educational inequality when AI platforms were deployed without equitable access measures. World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 903 Table 1 Adoption levels, reported benefits, and barriers across finance, healthcare, and education. SN. Sector Adoption/Im provement Reported Benefits Key Barriers Identified 1 Finance 62% gained first-time loan access Expanded credit access for unbanked, support for MSMEs Gender gap, smartphone access, algorithmic bias 2 Healthcare 70% health workers aided by AI tools Faster consultations, predictive maternal health apps Connectivity issues, dataset mismatch, privacy 3 Education 55% improved STEM scores Personalized learning, reduced teacher burden Cost barriers, skepticism, digital divide Figure 1 AI Adoption in Underserved Communities by Sector 5 Key Challenges The integration of Artificial Intelligence (AI) into information systems for inclusive innovation presents a paradoxical reality: while the technology carries immense potential to expand access to finance, healthcare, and education, it also introduces profound risks that may entrench inequality. A critical reflection on both challenges and opportunities is therefore necessary to understand how AI can be leveraged responsibly. This section synthesizes insights from focus group discussions (FGDs), key informant interviews (KIIs), surveys, and an extensive literature review to outline the barriers that currently hinder inclusivity and the transformative possibilities that AI creates when effectively implemented. 5.1 Algorithmic Bias and Discrimination One of the most significant challenges revealed in FGDs is algorithmic bias and discrimination. Participants in refugee settlements and rural communities often expressed concern that AI-driven financial services or educational platforms did not adequately reflect their realities. KIIs with technology developers confirmed that many algorithms used in Africa and Asia are trained on datasets from Western populations. As a result, systems frequently misinterpret behavioral World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 904 patterns that are culturally specific, leading to unfair outcomes. For example, mobile usage data that is considered “irregular” in credit scoring models might be perfectly normal in a context where mobile coverage is patchy and income streams are seasonal. Literature reviewed from Harvard Data Science Review and the ACM FAccT Conference further confirmed that bias is one of the most intractable problems in AI adoption. Unless addressed through localized datasets and continuous auditing, bias threatens to replicate structural inequalities under the veneer of technological neutrality. 5.2 Infrastructural Deficits A second challenge lies in infrastructural deficits, repeatedly highlighted across surveys and FGDs. In rural areas, communities noted that while AI-enabled applications exist, access to them is severely hampered by unreliable internet connectivity, intermittent electricity, and high costs of smartphones or data plans. KIIs with policymakers acknowledged that infrastructure remains the Achilles’ heel of digital innovation in the Global South. Without investment in affordable and stable digital infrastructure, AI systems risk deepening the digital divide by privileging urban and wealthier users. Literature review of ITU and World Bank reports confirmed that digital exclusion remains strongly correlated with poverty and geography, creating uneven landscapes of opportunity. 5.3 Trust and Governance Frameworks Another challenge relates to trust and governance frameworks. Communities surveyed expressed skepticism about how their personal data was being stored, shared, and used. In healthcare, patients were particularly concerned about privacy, fearing that sensitive medical records could be misused by governments or corporations. KIIs with regulators revealed that many countries lack comprehensive data protection laws, let alone specialized regulations for AI. This governance vacuum creates risks of surveillance, misuse of personal information, and opaque decision-making processes. As literature from the OECD and UNESCO stresses, trust is a precondition for widespread adoption of AI systems, and without clear governance, communities may reject otherwise beneficial innovations. 5.4 Digital Literacy and Human Capacity The fourth challenge lies in digital literacy and human capacity. FGDs revealed that even when AI tools are available, many community members lack the skills to use them effectively. Surveys showed that a significant number of users were unable to distinguish between automated chatbot responses and AI-driven health advice, creating risks of misinterpretation. KIIs with educators and civil society organizations underscored that digital literacy programs are essential if AI is to function inclusively. Literature review supported this finding, with multiple studies arguing that without foundational digital skills, AI risks exacerbating inequality by concentrating benefits among already digitally literate populations. 6 Opportunities Despite these challenges, the opportunities presented by AI are equally significant. In finance, AI-driven credit scoring and digital lending platforms present a leapfrogging opportunity for financial inclusion. FGDs with small business owners highlighted how access to microloans through mobile platforms had enabled them to expand businesses, pay school fees, and cope with emergencies. KIIs with FinTech entrepreneurs confirmed that AI models can dramatically reduce transaction costs, making it feasible to serve customers historically excluded by traditional banking. Literature review further suggests that AI-enabled finance could accelerate progress toward poverty reduction and women’s economic empowerment. 6.1 Scalable Diagnostic and Predictive Systems In healthcare, the opportunity lies in scalable diagnostic and predictive systems. Surveys showed high levels of appreciation among patients using AI-enabled health apps, particularly during the COVID-19 pandemic when physical access to clinics was limited. KIIs with healthcare professionals emphasized that AI tools can augment, rather than replace, overstretched medical staff by providing early warning signals, supporting triage, and predicting outbreaks. Literature from WHO further confirmed that AI holds potential for universal health coverage, particularly in resourceconstrained settings. World Journal of Advanced Research and Reviews, 2025, 27(03), 897–908 905 6.2 Personalized and Adaptive Learning Education presents opportunities for personalized and adaptive learning. FGDs with students indicated strong enthusiasm for AI tutors that provided additional support in mathematics and science. Teachers interviewed noted that AI reduced grading workloads, allowing them to focus more on direct engagement. Literature review from UNESCO underscored that adaptive learning platforms can tailor instruction to individual needs, improving outcomes in overcrowded classrooms. If scaled equitably, AI could help bridge educational gaps and empower the next generation with critical skills. 6.3 Resilience and Sustainability Beyond these sectors, AI offers cross-cutting opportunities for resilience and sustainability. KIIs with agricultural stakeholders highlighted the potential of AI for crop monitoring, pest detection, and climate adaptation. International organizations interviewed emphasized that AI could improve humanitarian logistics, food distribution, and disaster response, enabling faster and more efficient aid delivery. Literature from UNDP and WFP supports this, showing that AI can strengthen resilience in fragile and conflict-affected contexts. 7 Key Insights and Recommendations The analysis of AI-driven information systems in finance, healthcare, and education reveals both transformative opportunities and pressing challenges for inclusive innovation. When triangulated across focus group discussions, key informant interviews, surveys, and literature review, several key insights emerge. These insights do not present AI as a neutral or inherently benevolent tool; rather, they underscore that inclusivity in AI is the result of deliberate choices, frameworks, and collaborative action. The recommendations that follow are therefore not generic but anchored in the empirical realities encountered during this research, combined with critical reflection on global debates surrounding ethics, equity, and technological governance. One of the most salient insights is that AI’s potential for inclusion is directly tied to its design and contextualization. Focus group discussions highlighted that communities often encounter AI systems that do not reflect their cultural, linguistic, or socio-economic realities. For instance, rural populations struggled with mobile applications available only in English or French, leaving local language speakers disadvantaged. Similarly, women in refugee settlements expressed skepticism about AI-driven financial services that required smartphone ownership, given that men in households disproportionately controlled access to digital devices. Key informants in technology firms acknowledged these gaps, noting that most AI models used in the Global South were trained on datasets collected in the Global North. This misalignment underscores the critical insight that inclusivity cannot be an afterthought; it must be deliberately engineered into the very architecture of AI systems. A second insight relates to the interplay of infrastructure and opportunity. Surveys revealed that even where AI applications were available, poor internet connectivity, lack of electricity, and high device costs hindered adoption. In healthcare, for instance, diagnostic AI tools piloted in rural clinics were abandoned after months due to the absence of stable power supplies. In education, students reported being unable to access online AI tutors consistently due to weak internet coverage. Literature review reinforced this finding, with multiple studies emphasizing that digital infrastructure is the backbone of equitable AI adoption. Without investment in this foundational layer, AI systems risk exacerbating exclusion by disproportionately serving urban, wealthier, and digitally literate populations. A third insight is the importance of trust and governance in AI adoption. Across FGDs, participants repeatedly voiced concerns about data privacy, surveillance, and opaque decision-making. In financial services, users worried about how their personal data was being stored and whether it could be misused by governments or corporations. In healthcare, patients expressed fear that sensitive medical records could be shared without consent. KIIs with policymakers revealed that regulatory frameworks for AI remain underdeveloped in many countries, with few legal safeguards against misuse. Trust emerged as a central determinant of adoption: communities were more willing to use AI systems when backed by international organizations or reputable NGOs than when delivered solely by private companies. This points to the critical insight that governance frameworks covering data protection, algorithmic transparency, and accountability are not peripheral but central to inclusive innovation. A fourth insight is the necessity of capacity-building and digital literacy. The research revealed a paradox: AI systems are often marketed as simplifying access, yet their effective use still requires a baseline level of digital literacy that underserved communities may not possess. In FGDs, participants described struggling with basic app navigation, while