In our technology-driven society, access to and mastery of digital tools are essential for accessing information, obtaining services and communicating. We are hearing more and more about artificial intelligence (AI) and its applications, particularly in the fields of health, employment and education.
However, leading figures have recently called for a halt to the development of certain advanced artificial intelligence systems, in order to better manage innovation and prevent potential harms.(1) Indeed, these systems are fed by data pumped from the Internet, regardless of its veracity. This data therefore reflects society's implicit and explicit prejudices, which can create misinformation, and exacerbate inequalities and different forms of prejudices: such as racism, sexism... and ageism.
Ageism can lead to social isolation, loneliness, financial insecurity, reduced quality of life and premature death. The World Health Organization estimates that more than 6 million cases of depression worldwide are due to ageism.(2) Older adults use the Internet less than younger people, but the gap is narrowing.(3) Online services, social networks and sharing platforms are tools that enable more and more older adults to be informed, socially engaged, autonomous and connected with family, friends and the community.
Given the aging population, the question arises: are age-related biases encoded and amplified in artificial intelligence (AI) systems?
Ageism and the digital divide
A recent evidence synthesis examined this question.(4) The results once again reveal the digital divide between those who have access to digital technologies and those who do not. This creates a disparity in access and limits social and community participation.
Despite the fact that more and more people are using the Internet, older adults are less likely to have access to it due to psychological factors or physical barriers such as disabilities, for example. Another barrier is that digital interfaces are rarely designed for this clientele. Those interfaces often use small print, low-contrast colours, confusing technical terms and laborious registration or authentication procedures.
Bias in AI systems
AI quickly processes a lot of data to establish correlations, predict behaviours and solve problems. On some topics, there isn't enough data available on older adults to train AI systems or create useful applications tailored to this population.
The perceptions and psychomotor abilities of older adults are not always taken into account when designing and developing technological applications, contributing to a low rate of technology adoption.
Some designers seem to believe that the only digital applications that older adults need are those that enable them to manage their health. This bias fosters the idea that older adults are a homogeneous group with health problems who have no other needs, interests or aspirations. AI systems will use such biased data and representations to perpetuate this discrimination, widening the gap between what is offered to older adults and their actual needs.
Ethical and legal implications
There is no uniform global legal framework for AI governance. The development of laws and regulations poses challenges in relation to their application within and beyond our borders. In addition, in the Canadian context, healthcare and human rights are shared between the federal government and the provinces, making the exercise complex. Moreover, beyond jurisdictional issues, governments must ensure that they protect the public without stifling innovation, while being aware that some AI algorithms benefit from intellectual property protection.
The COVID-19 pandemic has accelerated the use and adoption of digital technologies and AI systems. More and more people will be integrating these into their lives, whether to access healthcare or to be socially connected. There are a number of measures we can take today to facilitate the inclusion of older adults:
- Taking into account the needs of older adults to create intuitive, easy-to-use interfaces adapted to various levels of technological skills.
- Offering more training and resources to develop older adults' technological skills.
By working together to overcome the challenges of digital ageism, we can create a digital world where every generation has a voice!