Opinion: How community-led innovation can help drive equitable AI

Artificial intelligence, or AI, is transforming our health care landscape — reshaping how people access and use health information and expanding what’s possible in diagnosis, treatment, and service delivery. But without intentional action, it risks reinforcing inequities that have long shaped global health.

As we explore the next frontier of tech innovation, we have a responsibility — and an opportunity — to ensure that its benefits are not reserved for people living in countries with the most resources. This is where philanthropy can play a vital role — guiding the deployment of AI in ways that advance equity from the start by capturing and sharing learnings on how AI can transform health delivery and be adapted to meet different needs in low-resource settings.

Many AI systems are developed using data that can exclude low- and middle-income countries, or LMICs, resulting in tools that don’t reflect the realities of the people they intend to serve. Today, many low-income countries don’t have the digital infrastructure to support AI platforms, and representative data remains limited. For these tools to make a positive impact, we must develop solutions that take these conditions into account and create protections to prevent the misuse of data and protect privacy.

At The Pfizer Foundation, we believe listening to communities and letting their insights guide innovation is essential to advancing health equity. That’s why we’re investing in locally driven solutions that apply AI not as a one-size-fits-all tool, but as a means to meet people where they are. Through our Global Health Innovation Grants, or GHIG, program, we’ve created an innovation incubator to support organizations that use AI to expand access in meaningful ways — demonstrating that when communities lead, technology can benefit everyone.

Since 2015, the GHIG program, managed in partnership with Innovations in Healthcare, has supported community-based organizations and initiatives that work to strengthen health systems and improve the quality of care in LMICs. To date, GHIG has collaborated with more than 45 organizations across 30 countries to help improve access to quality health services for more than 11.4 million people and established over 3,800 new points of care.

As technology and the health care landscape have evolved, so has the GHIG program. This year, we launched the AI Learning Lab, partnering with six organizations that are testing AI applications in diverse, real-world settings to address local health challenges.

Through the AI Learning Lab, we’re supporting community-led organizations as they apply new technologies in ways that are effective, inclusive, and locally relevant — from bringing critical care closer to new mothers to strengthening vaccine delivery. Real-world insights from community-led approaches like these are key to shaping equitable AI.

Here are three examples of how our network is putting that potential into action to help close the health equity gap.

Connecting mothers to timely maternal care

Ghana has made significant progress in reducing infant and maternal deaths over the last decade. However, many expectant and new mothers struggle to navigate care, recognize risk signs in pregnancy, and connect with timely maternity care — challenges that can lead to preventable complications. 

Jacaranda Health, a long-standing GHIG partner, is a Kenyan social enterprise dedicated to improving maternal and newborn health in underserved communities through technology and data-driven solutions. Building on the success of Jacaranda Health’s PROMPTS platform in Kenya, the AI-enabled SMS-based service is now expanding to Ghana to deliver personalized two-way communication to support maternal health, in partnership with the Ghana Health Service.

More than 3 million mothers have enrolled in PROMPTS, which encourages early care-seeking behavior, reinforces maternal health knowledge, and helps women recognize and report danger signs during pregnancy and postpartum. Locally contextualized large language models, or LLMs, help the platform identify and escalate urgent maternal health concerns, personalize messages for high-risk mothers in local languages, and improve help-desk efficiency with faster connections to lifesaving care.

Strengthening vaccine delivery through digital identity

Around 850 million people across the globe are unregistered and without formal identification, making it difficult for governments and health systems to deliver vaccines and track care over time.

Simprints, a nonprofit technology company that builds ethical, inclusive digital ID systems to help ensure lifesaving services reach underserved populations, is addressing this challenge with AI-powered biometric tools, including fingerprint and facial recognition. Simprints is strengthening malaria vaccination efforts in Ghana by helping ensure patients are accurately identified and followed up with throughout the vaccination cycle.

By using AI to help eliminate duplicate records, improve data quality, and ease the administrative burden, Simprints supports health workers in delivering services more efficiently. Their biometric technology can help reduce the average time spent per clinic visit and enable real-time identification of children who missed vaccine doses. These tools are supported by strong privacy safeguards and community engagement strategies, which are essential for building trust and ensuring ethical, inclusive implementation. To date, the project has created over 65,000 unique electronic medical records, supporting better coverage and helping close critical gaps in access to lifesaving vaccines.  

Promoting early detection through community-based AI tools

In Colombia, cervical cancer is the leading cause of cancer-related deaths among women and a significant public health concern. Fundación Vive con Bienestar, or Bive, is an organization that connects at-risk rural families to fast, affordable, high-quality health care. Bive is implementing its Protégete con Bive2, or PCB2, model to expand prevention and early detection of human papillomavirus, or HPV — a common viral infection that increases women’s risk of contracting cervical cancer.

The model combines two AI-powered innovations: Pegasi, which uses predictive algorithms to help identify high-risk cases early, and JoyChatbot, a youth-friendly tool that taps into natural language processing to share evidence-based information on sexual and reproductive health. Deployed through both clinical settings and community outreach, these solutions — alongside Bive’s network of community health workers — aim to reduce cervical cancer mortality and strengthen care infrastructure where it’s needed most.

These programs show that when AI is guided by those closest to the challenges, it can help reduce delays in care, foster trust in health systems, and strengthen more equitable models of delivery. By investing in locally-led solutions, we make space for new, community-centered approaches that can transform health outcomes and build stronger, more inclusive health systems.

Philanthropy plays a powerful role in this journey — providing early support to promising ideas and organizations working toward system-level change. As global health funding becomes increasingly constrained, philanthropic investment offers a vital pathway to support exploration and innovation through new technologies and approaches that can grow, scale, and achieve lasting impact.

Looking ahead, we must continue investing to better understand what works, share learnings across settings, and adapt approaches for the long term. When local leadership shapes decision-making from the start, equity becomes foundational — and innovation can deliver lasting change.

The Pfizer Foundation is a charitable organization established by Pfizer Inc. It is a separate legal entity from Pfizer Inc. with distinct legal restrictions.