Introduction

Artificial intelligence (AI) is increasingly finding its place in healthcare, and experts believe it could transform maternal and newborn health by helping expectant mothers receive timely care while supporting clinicians in making faster, informed decisions. However, they caution that the technology must complement—not replace—medical professionals and should be governed by strong ethical, legal and data protection frameworks.

AI to Reduce Delays in Healthcare

Medical doctor and digital health expert, Dr, Victor Barasa says AI presents an opportunity to convert existing medical knowledge, including clinical guidelines, protocols and best practices, into automated tools that can deliver healthcare information more efficiently and at a lower cost. “Artificial intelligence allows for early triaging and decision-making,” he explains. “It addresses the three major delays in maternal healthcarethe delay in deciding to seek care, the delay in reaching a health facility and the delay in receiving an accurate diagnosis.” These three delays remain among the leading contributors to maternal deaths in many low- and middle-income countries, including Kenya. By identifying pregnancy risks earlier and prompting women to seek medical attention before complications worsen, AI could help reduce preventable maternal and newborn deaths. One example is PROMPTS, an AI-supported two-way SMS platform developed by Jacaranda Health. The service sends expectant mothers tailored health information, appointment reminders and guidance throughout pregnancy. According to Dr. Barasa, the lifetime cost of providing the service to one mother is less than one US dollar, making it an affordable solution for resource-constrained settings. He notes that the use of SMS technology makes the platform practical in Kenya, where smartphone ownership and internet connectivity remain uneven, particularly in rural communities.

The Limits and Risks of AI

Despite its potential, Dr. Barasa warns that AI should never replace clinical judgment. Instead, it should enhance healthcare delivery by supporting doctors, nurses and midwives in making better decisions. “There is a risk of over-reliance on AI instead of scientific methods and clinical skills,” he says. “Healthcare providers must continue to apply independent clinical judgment because AI systems can make mistakes.” He also raises concerns about patient privacy and data security, noting that AI platforms often process sensitive personal information, including names, locations, identification details and medical records. If not adequately protected, such data could be vulnerable to misuse. Dr. Barrazo says Kenya has already taken important steps to safeguard health data through the Digital Health Act, 2023, and the Office of the Data Protection Commissioner. He also welcomes the ongoing development of Kenya’s Artificial Intelligence and Emerging Technologies Policy but argues that regulations should encourage responsible innovation rather than create unnecessary barriers.

Ensuring Reliable and Responsible AI

Instead of rigid regulation, he advocates for regulatory “sandboxes,” where innovators and regulators can work together to safely test emerging technologies before they are deployed at scale. To ensure AI systems are safe and reliable, Dr. Barasa recommends rigorous local validation using Kenyan health data. He says multidisciplinary teams including obstetricians, midwives, pharmacists, data scientists and patients should be involved in evaluating AI tools to eliminate bias and ensure they align with Kenya’s clinical guidelines. Accountability, he adds, remains a critical issue. If an AI system fails to detect a danger sign during pregnancy, clinicians are still expected to verify its recommendations before making treatment decisions. “The doctor cannot simply rely on AI. Independent clinical judgment remains essential,” he says, adding that hospitals, software developers and healthcare providers must have clear accountability structures in place.

Conclusion

Ultimately, Dr. Barasa believes AI’s success should be measured by tangible improvements in maternal health indicators, including earlier antenatal care attendance, increased skilled deliveries, timely postnatal care and reduced delays in accessing healthcare. If implemented responsibly, he says, artificial intelligence could become a valuable partner in improving maternal and newborn outcomes while ensuring that patients remain at the centre of healthcare.