The World Health Organization recently released a document with six areas of focus for the development of artificial intelligence in healthcare.
What happened
WHO created a joint initiative alongside the International Telecommunication Union to establish a Focus Group on AI for Health (FG-AI4H). The group includes regulatory authorities, policymakers, and industry experts.
FG-AI4H recently published a document that provides an overview of regulatory considerations. The authors state that the publication is "not intended as guidance or as a regulatory framework or policy." Instead, it's designed to be a discussion and resource for those developing AI systems, as well as other stakeholders.
Going deeper
The 80-page document outlined six main focus areas for regulation of AI in the healthcare field:
- Transparency: FG-AI4H stresses the importance of transparency and documentation to improve public trust and understanding. They recommend documenting a product's lifecycle and tracking the development of AI systems.
- Risk management: The document suggests that organizations create simple models to address threats related to cybersecurity and issues related to training models, continuous learning, and more.
- Externally validating data: Organizations should also consider validating data externally and be clear about the intended use of AI. FG-AI4H believes this can help ensure safety and regulation.
- Data quality: Organizations should prioritize having high-quality data by evaluating every system before release. This also requires organizations to check for bias and errors within AI systems.
- Privacy protection: The document acknowledges the importance of following HIPAA regulations in the US and other regulations around the world that serve to protect patient privacy.
- Collaboration: Finally, the publication emphasizes the importance of AI developers coordinating with healthcare professionals, government partners, and other stakeholders when creating safe and effective products.
The document asserts proper regulation as a necessity, especially as it relates to preventing errors and bias in AI learning models. FG-AI4H hopes that governments can utilize this information to create new guidance or regulatory procedures applicable to their respective existing policies.
Related: HIPAA Compliant Email: The Definitive Guide
What was said
Director-General of WHO, Dr. Tedros Adhanom Ghebreyesus, said in a press release, "Artificial intelligence holds great promise for health, but also comes with serious challenges, including unethical data collection, cybersecurity threats and amplifying biases or misinformation."
The Director-General further said, "This new guidance will support countries to regulate AI effectively, to harness its potential, whether in treating cancer or detecting tuberculosis while minimizing the risks."
The big picture
While there are many possible benefits, developers must tread carefully to ensure patient privacy and effective emerging technologies. Leaders in the field suggest that AI development should be slow and steady, with a primary focus on providing value to patients.
Read more: "We're not in a hurry," Google Research Expert discusses responsible use of AI.
Abby Grifno
