Across U.S. health systems, Black leaders are showing how artificial intelligence and data science can narrow inequities—when these tools are designed and governed with equity at the center.
Dr. Kemi Doll (University of Washington) has helped expose how “off-the-shelf” algorithms and protocols can fail Black women in gynecologic care. Her team’s research—reported from a JAMA Oncology study—found that a common triage approach for endometrial cancer (relying on transvaginal ultrasound thresholds) can miss disease in Black women, underscoring the need for more equitable, evidence-based decision pathways and model validation across populations. Beyond that single study, Doll’s broader body of work documents the mortality gap and structural drivers of delay in care—groundwork that informs how AI tools should be evaluated for disparate impact before deployment.
Dr. Irene Dankwa-Mullan (formerly IBM Watson Health, now Merative) has been a prominent voice on building inclusive, human-centered AI—from data collection through deployment. Her federal advisory presentation emphasized integrating health-equity principles into AI/ML design and implementation, a practical blueprint for health systems procuring or building models. Her role leading health-equity strategy at IBM Watson Health/Merative further grounds her advocacy in real-world enterprise AI programs.
Dr. Elaine O. Nsoesie (Boston University) applies machine learning and computational social science to understand and address racial health inequities—showing how nontraditional data (e.g., social media signals) can inform targeted interventions in communities with fewer resources. She also argues for a “Data Science for Racial Health Equity” that explicitly includes social determinants of health—an essential design guardrail for any AI used in public health. Her lab’s published ML studies illustrate these methods in practice.
Dr. Toyin Ajayi, CEO and co-founder of Cityblock Health, is scaling value-based care for Medicaid and dually eligible members with data-driven models that help multidisciplinary teams prioritize outreach and personalize care—an operational cousin to AI that depends on similar pipelines, governance, and continuous learning. Cityblock’s approach—integrating medical, behavioral, and social care—demonstrates how analytics can direct scarce resources toward those most at risk.
Finally, Dr. Karriem S. Watson, chief engagement officer for the NIH All of Us Research Program, has led the creation of one of the world’s most diverse biomedical datasets—critical infrastructure for fairer AI/ML in precision medicine. All of Us operates a secure cloud Researcher Workbench that enables advanced analysis; NIH highlights how pairing AI methods with this diverse data can accelerate discoveries, while AIM-AHEAD training programs explicitly use All of Us to grow AI/ML capacity among diverse researchers.
Why this matters for 2025: Each leader illustrates a pillar of FLITE. Foresight: identifying where algorithms can harm and setting stricter validation standards (Doll). Leadership Readiness: embedding equity into AI governance frameworks (Dankwa-Mullan). Innovation: novel ML methods and equity-aware data curation (Nsoesie). Team Agility: operationalizing analytics across care teams (Ajayi). Emerging Technology: building diversified data platforms that make fair AI possible.
The lesson is clear: AI can widen gaps if built on biased data and weak oversight—or it can help close them when equity leaders shape the questions, the data, and the deployment.
Marvin Dejean is the CEO and senior managing partner of Gilead Sanders LLC, a strategy and business transformation consulting practice based in Fort Lauderdale. He empowers organizations to navigate and succeed in the digital age by leveraging future-focused strategies and cutting-edge technologies. He can be reached at mdejean@gileadsanders.com or online at www.gileadsanders.com.
