WHO: It’s Time to Eliminate Ageism in Artificial Intelligence

As caregivers examine how artificial intelligence (AI) can improve healthcare for seniors, a policy brief by the World Health Organization shows that potential issues can be eliminated by taking various steps, including optimal design and data collection, providing resources, and maintaining good moral codes.

The most common uses of healthcare-focused AI for seniors are remote patient monitoring, which is used as a method of community care, and the production of drugs for aging patients. The report, released Wednesday, also stated that optimizing AI capabilities can increase hospital capacity, as beds can be reserved for cases requiring hands-on care.

The risks associated with AI-driven healthcare revolve mainly around ageism, which is discrimination based on one’s age that can affect how useful a resource is.

New AI approaches require data utilization, but older people are often overlooked for data collection. Data sets containing skewed information are frequent, whether intentional or not.

Historically, there have also been issues relating to an uneven distribution of technological services. Older people may be discouraged from using AI due to the fixed belief that they are not as efficient with technology.  

Further, because creators do not often consider the older generation when designing novel technology, biased perspectives on AI use emerge.

Feedback of AI also often neglects the voices of older people, which causes a struggle in maintaining applicability. While creating guidelines, creators need to confirm that the elderly receive proper care and attention.

To maximize benefits and avoid pitfalls, the report detailed a list of eight considerations that must be present when creating and implementing healthcare AI technology for seniors.

First, the design of services should include the people using them, in this case, older people. Creating a well-balanced and diverse group in all demographics can optimize quality and prevent bias.

In addition, people of various ages should be included in data samples. If the collected selection is not diverse, the benefits will only extend to a small age range.

There is a need for a digital infrastructure that promotes a greater understanding of how to use AI technology among seniors. Older people must be kept well informed about evolving technology.

AI services must also maintain fail-safety, that is, they must be designed to ensure that errors do not result in adverse events. First, all users must have the option of withdrawing from any AI-driven procedure. Second, existing regulations that eliminate ageist perspectives to avoid negative consequences must be enacted.

Finally, there must be an emphasis on promoting research and practicing ethics. It is essential that creators of AI learn from errors and engage with research on how strategies and codes of ethics grow.

"The implicit and explicit biases of society, including around age, are often replicated in AI technologies," said Alana Officer, unit head of Demographic Change and Healthy Ageing at WHO, in a post online. "To ensure that AI technologies play a beneficial role, ageism must be identified and eliminated from their design, development, use and evaluation."