WHO has formulated 'Six Basic Philosophy for Using AI for Healthcare', what is its content?

On June 28, 2021, WHO announced 'Six Basic Principles for Making Artificial Intelligence (AI) Work for the Public Interest in All Countries,' which will guide WHO's future activities. did. In it, WHO argues that AI should not violate human freedom and human rights, and that it should be able to benefit everyone in the world.

WHO issues first global report on Artificial Intelligence (AI) in health and six guiding principles for its design and use

Artificial Intelligence ( #AI ) holds enormous potential for improving the health of millions of people around ????????????, but only if ethics & human rights are put at the heart of its design, deployment, & use.

More in WHO's first global report on AI & Health ???? https://t.co/fC7SX51YzY pic.twitter.com/1Jp0ASPDY3

— World Health Organization (WHO) (@WHO) June 28, 2021

On June 28, WHO released '(PDF file) Ethics and Governance of Healthcare AI ', a report summarizing the results of two years of discussions between international experts on AI and healthcare. did.

In a statement, WHO Executive Director Tedros Adhanom Gebreyesos said, 'Alike other new technologies, AI has great potential to improve the health of many people around the world, but it is harmful due to misuse. The important report released this time is to avoid the traps hidden in the use of AI, minimize the risk of using AI, and maximize the benefits. It shows the world what to do. '

The 'Basic Philosophy for Healthcare AI' proposed by WHO in the report is the following six points.

◆ 1: Protecting human autonomy
Autonomy here means that humans control healthcare systems and medical decisions, ensure privacy and confidentiality, and provide appropriate informed consent to patients through appropriate legal systems for data protection. It means that you will be given an opportunity.

◆ 2: Promoting human well-being and safety and the public interest
When designing AI technology, you must meet well-defined use case safety, accuracy, and effectiveness criteria. In addition, for practical use, means for quality control and quality improvement must be prepared.

◆ 3: Ensuring transparency, explainability, and comprehension
To ensure AI transparency, sufficient information must be published in writing before AI is designed. This information should also be easily accessible to everyone and useful for public discussions about what AI should be.

◆ 4: Fostering responsiveness (responsibility) and accountability (accountability)
AI stakeholders are responsible for ensuring that AI is used by the right trained people under the right conditions. Appropriate explanations and remedies must also be given to those affected by AI-based decisions.

◆ 5: Ensuring inclusion and fairness
Inclusion is a concept that expresses age, gender, income, race, ethnicity, sexual orientation, ability, etc. Healthcare AI must be made to be used and used fairly regardless of these personal characteristics.

◆ 6: Promotion of responsive and sustainable AI
Everyone who designs, develops, and uses AI needs to continually evaluate its capabilities while using it to determine if it meets expectations and requirements. AI should also be energy efficient to minimize its impact on the environment. In addition, governments and businesses must properly train health care workers who use AI and reduce the disruption to those who lose jobs and the workplace due to AI.

WHO has positioned this basic philosophy as a guideline for WHO's activities in the future, and based on this, will work to make healthcare AI in the interests of all.