The World Health Organization (WHO) has issued guidelines for regulating artificial intelligence (AI) in healthcare, emphasizing safety, effectiveness, and stakeholder collaboration. It recognizes AI’s potential in transforming healthcare, improving clinical trials, diagnostics, treatment, and supplementing medical knowledge. Despite the rapid deployment of AI technologies, concerns arise about performance, ethical considerations, and potential harm to end-users. The WHO underscores the need for robust legal and regulatory frameworks to protect privacy, security, and integrity in handling sensitive health data.
WHO Director-General Dr Tedros Adhanom Ghebreyesus states that while AI holds great promise for health, it also poses challenges like unethical data collection, cybersecurity threats, and biases. The new guidance aims to assist countries in regulating AI effectively, balancing its potential benefits with minimized risks.
The publication outlines six key areas for AI health technology regulation:
Transparency and Documentation:
• Stresses transparency throughout the AI product lifecycle.
• Emphasizes documenting development processes to foster trust.
Risk Management:
• Addresses issues like ‘intended use,’ ‘continuous learning,’ human interventions, and cybersecurity threats.
• Advocates for simplifying AI models for effective risk management.
External Validation:
• Highlights the significance of externally validating data for safety and regulation facilitation.
• Calls for clarity on the intended use of AI systems.
Data Quality Commitment:
• Advocates for a commitment to data quality through rigorous pre-release evaluation.
• Vital for preventing biases and errors in AI systems.
Compliance with Regulations:
• Addresses challenges from complex regulations like GDPR and HIPAA.
• Emphasizes understanding jurisdiction and consent requirements for privacy and data protection.
Collaboration for Lifecycles:
• Encourages collaboration among regulatory bodies, patients, healthcare professionals, industry representatives, and government partners.
• Aims to ensure compliance with regulations throughout the lifecycles of AI products and services.
The WHO publication aims to provide governments and regulatory authorities with key principles for developing or adapting AI guidance at national or regional levels.
For details, please click here.