The World Health Organization (WHO) published the first global report on the use of artificial intelligence in medicine and outlined the core principles for its design and use.
According to the report, AI holds great promise for improving medical care and health systems worldwide. Yet ethics and human rights must underpin its development and use.
“Like all new technologies, artificial intelligence holds enormous potential to improve the health of millions of people around the world, but it can also be misused and cause harm,” said WHO Director-General Tedros Adhanom Ghebreyesus.
In some developed countries, AI is used to accelerate and improve the accuracy of diagnosing diseases, deliver clinical care, advance health research and drug development, and support various public health measures, the report says.
However, researchers warn that deploying the technology carries risks—from unethical collection and use of health data, to algorithmic biases and patient safety concerns. They also do not rule out potential threats to the environment and cybersecurity.
To reduce risks in the development and use of AI, researchers proposed six principles aimed at ensuring safety of the technology in health care:
- Protecting human autonomy — people should retain control over health systems and medical decisions, and expect privacy.
- Promoting the well-being and safety of people and the public interest — developers should guarantee the safety, accuracy and effectiveness of the technologies.
- Ensuring transparency, explainability and understandability — developers of AI systems should publish sufficient information for public consultation.
- Increasing accountability and responsibility — stakeholders should bear responsibility for the misuse of technologies.
- Ensuring inclusivity and fairness — equitable use and access to medical AI, regardless of social group.
- Advancing flexible and resilient AI — continuous and transparent evaluation of the technology in use to assess its alignment with expectations and requirements.
Earlier in June, UNESCO researchers concluded that in the last five years artificial intelligence and robotics have become dominant scientific fields.
In May, neural networks and brain implants helped a paralysed man write text messages.
At the beginning of May, a group of American researchers concluded that patients do not trust AI doctors, who know their names.
Subscribe to ForkLog on Telegram: ForkLog AI — all the news from the world of AI!