Publication: Platforms for artificial intelligence-enabled infectious disease surveillance (npj Digital Medicine, 2026)

In recent years, the rapid evolution of artificial intelligence (AI) has sparked a surge in AI-powered infectious disease surveillance tools. As options proliferate, public health leaders, practitioners, and researchers face a growing challenge in distinguishing between platforms and choosing the right tool for the right job.

In a paper recently published in npj Digital Medicine, Marya Getchell and colleagues reviewed publicly available AI-enabled surveillance platforms focused on human health. Utilising a standardised framework, platforms were characterised according to their primary data inputs, surveillance function, AI methods, data privacy, geographic scope, and pathogen focus. The twenty platforms that met the inclusion criteria clustered into three functional archetypes:

• Early warning networks – Broad horizon scanning
  • Primary objective: Scour massive volumes of unstructured, open-source data (e.g. news feeds, social media, and academic publications) for rapid, early signal detection.
  • Strengths and limitations: High global speed and breadth, but limited capacity for localized risk interpretation or targeted action.
  
  
• Situational awareness platforms – Regional context and outbreak forecasting
  • Primary objective: Provide localised assessments of specific pathogen risks by integrating epidemiological data with climate, demographic, and mobility variables, often utilising restricted national health datasets.
  • Strengths and limitations: Highly actionable for national or regional policy, but heavily dependent on local data quality and infrastructure sustainability.
  
  
• Integrated surveillance platforms– Targeted public health decision support
  • Primary objective: Link molecular pathogen data to broader population-level transmission and environmental context to deliver high-resolution outbreak characterisation and provide public health decision-making support.
  • Strengths and limitations: Provides precise analysis of transmission dynamics, outbreak risk, and intervention needs, but faces high implementation demands and complex interoperability and governance challenges.

Rather than competing solutions, these three archetypes represent complementary layers of a modern infectious disease surveillance system. By understanding where each tool excels, public health agencies can move away from fragmented technology adoption and build coordinated and comprehensive surveillance ecosystems.

Figure 1. Functional archetypes of AI-enabled infectious disease surveillance platforms. Platforms are grouped according to their function along the infectious disease surveillance continuum, indicating depth of support for decision making, and are also grouped according to variation in data inputs to indicate degree of data integration. Early Warning Networks (N = 10) primarily use open text-based data for signal detection; Situational Awareness Platforms (N = 4) integrate epidemiologic and geospatial data for localised risk assessment and disease forecasting; and Integrated Surveillance Platforms (N = 6) combine genomic, epidemiologic, and contextual data for higher-resolution interpretation and decision support. Bioinformatics tools were excluded from detailed comparative analysis but are shown here to illustrate their use of biological data and foundational support toward public health decision making. BLAST: Basic Local Alignment Search Tool.