Generational divide defines growing public anxiety over AI bias

Research from HyperFinity underscores that the perception of bias is far from uniform across age groups. While 51% of the general population acknowledges the presence of gender bias in AI, this figure climbs to 70% among 18-24 year olds. In contrast, only 37% of those aged 55-65 express similar concerns. This younger demographic is not merely more skeptical; they are also the most frequent users of the technology, leading to more direct and personal encounters with algorithmic limitations.

Public perception of these tools often relies on arbitrary, gendered associations. Users frequently project human traits onto software, with 46% of adults describing Siri as female and 35% associating ChatGPT with a male persona. These perceptions influence how users view the utility of the tools: ChatGPT is commonly framed as an expert advisor, whereas Alexa is perceived by 69% of respondents as female and is disproportionately labeled as a servant.

These biases carry real-world weight, particularly in critical sectors like healthcare and employment. With 43% of UK adults turning to AI for medical advice, concerns regarding inequity are mounting. Jessie Stuart, Head of Analytics at HyperFinity, warns that these issues are structural rather than accidental. Because AI systems are trained on historical data, they often encode decades of societal inequality into their foundations. Stuart cautions that without active intervention, society risks institutionalizing these biases in the very systems now used to make vital life decisions.