WHO: Careful use of artificial intelligence in healthcare

The World Health Organization (WHO) warned Tuesday that special care should be taken when using artificial intelligence (AI) in health care because of the risk of bias or misuse of data used for decision-making.

There is no doubt that AI applications are rapidly gaining popularity and demonstrating a trend that could disrupt the way businesses and society operate, and certainly has the potential to revolutionize the healthcare sector.

Artificial intelligence can be, and is already being used in some wealthy countries to improve the speed and accuracy of disease diagnosis and screening; assist in clinical care; enhance health research and drug development, and support diverse public health interventions such as disease surveillance, outbreak response, and health system management.

WHO notes that there is a great deal of anticipation about the potential of AI and the promise of its applications, but there are still concerns about how it will be used in practice to enhance access to medical information, as a decision support tool and to improve diagnosis.

As in other industries, the medical data used to train AI can be biased and produce misleading or inaccurate diagnostic information, and even these models can be misused to produce false information.

The WHO says assessing the risks of using large language models (LLMs) of artificial intelligence such as ChatGPT is "necessary" to protect and promote human well-being and protect public health.

The organization noted that the hasty adoption of untested AI systems could lead to errors by healthcare professionals, cause harm to patients and erode trust in AI, thereby undermining or delaying the potential long-term benefits and use of such technologies globally.

While working to improve human health using new technologies such as AI and digital health, WHO recommends that policymakers prioritize ensuring patient safety and protection.

WHO emphasizes six principles to ensure that AI is in the public interest of all countries: i. protect human autonomy; ii. promote human well-being, safety, and the public interest; iii. ensure transparency, interpretability, and comprehensibility; iv. promote responsibility and accountability; v. ensure inclusiveness and fairness; and vi. promote responsible and sustainable AI.

At the same time, doctors and public health experts in multiple countries have called for the development of general AI to be halted until it is regulated. They see risks associated with medicine and healthcare including the potential for patient injury from AI errors, data privacy and security concerns, and the use of AI in ways that exacerbate social and health inequalities.