INTRODUCTION:
Artificial intelligence (AI) offers new opportunities to generate timely, actionable insights from routine health data, particularly in low- and middle-income countries where such data are underutilized. When integrated into digital health systems, AI-driven tools have the potential to strengthen health system performance by improving planning, targeting, and responsiveness. This panel, led by the Government of Meghalaya, examines how AI-enabled analytics and feedback mechanisms are being applied in South Africa and India to enhance the use of routine health data.

METHODS:
The panel draws on comparative experiences from South Africa and the Indian state of Meghalaya. In South Africa, the analysis focuses on the application of AI techniques—including predictive modelling, natural language processing, and automated data processing—within a mature health data exchange built on long-term investments in patient registration systems, unique identifiers, and interoperable digital platforms. In Meghalaya, the discussion centres on a growing ecosystem of digital health applications for pregnancy tracking, immunisation, and non-communicable disease screening, which capture real-time data from frontline health workers, facilities, and programme units. Across both contexts, the panel examines the integration of AI-driven predictive analytics, automated data quality checks, and digital feedback mechanisms into routine health workflows.

RESULTS:
In both settings, AI-enabled approaches show promise in improving service delivery planning, identifying gaps in continuity of care, detecting high-risk cases, forecasting service utilisation, and strengthening quality assurance through anomaly detection and data completeness monitoring. Embedding AI-generated insights into dashboards, alerts, and supervisory tools enables more timely and targeted action by frontline workers, managers, and community health structures.

CONCLUSIONS:
AI can play a transformative role in strengthening health systems when embedded within practical, user-centred feedback loops. The panel highlights key considerations for responsible implementation, including infrastructure and workforce readiness, data governance, and safeguards to ensure that AI enhances decision-making without increasing workload or reinforcing inequities.