Artificial Intelligence Tools for Public Health

The adoption of artificial intelligence in public health requires assessing institutional capacities, governance, infrastructure, and readiness for its responsible implementation.

The PAHO Readiness Assessment Tool for Artificial Intelligence in Public Health helps identify strengths, gaps, and opportunities to support the adoption of artificial intelligence within health systems.

The approach proposes a comprehensive assessment that considers aspects related to governance, digital infrastructure, technical capacities, data management, regulation, ethics, and institutional readiness for the responsible use of emerging technologies.

AI-GUARD Tool. Governance & Use Assessment for Responsible Deployment

Download

Artificial Intelligence in Public Health: Readiness Assessment Toolkit

Download

NEW

AI-GUARD Tool: Governance and Use Assessment for Responsible Deployment

AI-GUARD is a tool developed by PAHO and the IDB to help decision-makers at all levels of health institutions assess the readiness and risk profile of any AI initiative before its adoption, procurement, development or scale-up.

AI-GUARD is a structured, tiered instrument designed to:

  • Assess the strategic public health value of AI initiatives.
  • Identify risk exposure and potential impact.
  • Assess institutional governance and operational readiness.
  • Identify and mitigate bias and equity-related concerns.
  • Strengthen transparency, accountability and responsible deployment.
     

The tool promotes disciplined, evidence-based decision-making while enabling innovation aligned with public health priorities.

The Four Core Pillars of AI-GUARD
 

1

Governance and accountability

Assesses the definition of responsibilities, oversight mechanisms, incident management and transparency in updates to artificial intelligence solutions.

2

Human use and control

Considers human intervention, override capability, user training, integration into workflows and accountability for final decisions.

3

Safeguards against risks and bias

Examines data representativeness, performance across different subgroups, the use of proxy variables, equity metrics and continuous monitoring to identify and mitigate bias.

4

Implementation readiness

Reviews the robustness of testing, validation status, monitoring plans, drift detection and the sustainability of the solution over time.

Artificial Intelligence in Public Health: Readiness Assessment Toolkit

This toolkit provides a structured approach to assessing a country’s readiness to implement AI projects in public health. Each question is designed to facilitate discussion and planning by identifying strengths and areas for improvement.
 

What does the toolkit assess?

The toolkit makes it possible to analyse different dimensions that influence the readiness of countries and institutions to adopt artificial intelligence in public health in a responsible, safe and sustainable manner.

Governance and leadership

Assesses institutional capacity to define priorities, coordinate stakeholders, and guide policies related to artificial intelligence in health.

Infrastructure and technology

Considers the availability of digital infrastructure, connectivity, interoperability, and technological capacities needed to implement AI solutions.

Data and information management

Analyzes the quality, availability, protection, and governance of the data required for the development and use of artificial intelligence in health.

Capabilities and human talent

Identifies technical skills, professional profiles, and institutional capacities needed to design, implement, and oversee AI solutions.

Ethics and regulation

Reviews the existence of regulatory frameworks, transparency mechanisms, data protection, equity, and human oversight.

Implementation and sustainability

Assesses the conditions required to integrate artificial intelligence solutions into real health system processes and sustain them over time.

How to use the toolkit?

The tool is designed to support institutional assessment and planning processes related to the adoption of artificial intelligence in public health. Its application helps identify existing capacities, recognize gaps, and prioritize strengthening actions.

1

Review dimensions

Analyze the areas included in the assessment and the associated criteria.

2

Complete assessment

Identify institutional strengths and challenges related to AI.

3

Identify gaps

Recognize priority areas to strengthen capacities and processes.

4

Prioritize actions

Define lines of work to advance readiness and governance.

5

Plan implementation

Guide sustainable processes for integrating AI into public health.

Readiness levels

The tool helps identify different levels of institutional readiness to adopt artificial intelligence in public health, supporting action planning and the progressive strengthening of capacities.

1

Initial

Institutional and technological capacities are limited. Challenges exist in governance, infrastructure, data, and human resources.

2

Developing

Partial progress is observed in digital transformation, data management, and the definition of institutional AI frameworks.

3

Advanced

The institution has more consolidated capacities in infrastructure, governance, and the strategic use of data for AI solutions.

4

Consolidated

Artificial intelligence is sustainably integrated into public health processes through ethical, interoperable, and people-centered approaches.