The Olympic Games continue to contribute to the conversation about stereotypes in sports. Particularly, those surrounding the physical appearance of certain athletes. However, a recent study conducted by Edith Cowan University has delved into the discussion by sharing the outcomes of their research, on the connection between the Olympics and AI. The findings unveiled how Artificial Intelligence perceives the teams of nations and brought to light the biases of AI-generated visuals – spotlighting the importance of scrutinising AI results for fairness and accuracy.
The research examined data from 40 countries using cues like “the Olympics team ” “Team GB,” and “photo” to create images of each nation’s team. The analysis revealed biases such, as gender, event type, cultural backgrounds, historical influences and emotional aspects drawing attention to the stereotypes ingrained in AI systems.
Gender bias
Of the images generated, 82.86% depicted men, and only 17.14% depicted women. Men appeared in images nearly five times more often than women. This finding isn’t just about AI. It’s a reflection of a broader issue in sports where female athletes often receive less media coverage and recognition than their male counterparts. From lesser prize money to fewer sponsorship deals, gender bias in sports is a longstanding issue that AI now mirrors.
Event bias
The study also uncovered notable event biases. Depictions primarily portrayed Canada’s team as ice hockey players, Argentina’s team as footballers, and the Netherlands’ team as cyclists. This suggests a tendency for the AI to stereotype countries by their most internationally recognised sports. Disregarding the diversity of athletic events each nation’s Olympic team actually participates in.
Such biases echo common stereotypes in sports. While Canada is famous for ice hockey and Argentina for football, these countries excel in numerous other sports. However, the tendency to pigeon hole them into a single sport diminishes their athletes’ diverse talents and accomplishments.
Cultural bias
The AI relied on cultural stereotypes in its representation of certain teams. Depictions portrayed the Australian team with kangaroo bodies and koala heads, showed Nigeria’s team in traditional attire, and dressed Japan’s team in kimonos. These images suggest that the AI favoured cultural cliches over accurate representations of modern athletes from these countries.
This kind of bias isn’t new; cultural stereotypes have long influenced perceptions in sports. Athletes often face expectations based on their nationality, which overshadows their individual identities and achievements.
Historical bias
Depictions showed the Greek team wearing ancient armour, while the Egyptian team wore attire associated with ancient Egypt. Such as gilded neckpieces. These portrayals rely on historical imagery instead of representing modern athletes, reducing these countries’ rich contemporary cultures to outdated historical symbols. This bias overlooks the present day identities of these nations and perpetuates anachronistic stereotypes.
Historical biases can undermine the progress and modern identities of these nations. And, trap them in a narrative that ignores their current realities and achievements.
Emotional bias
The study also revealed instances of emotional stereotyping. The South Korean and Chinese Olympic teams were depicted with stern expressions, reinforcing the stereotype that Asians are more reserved and serious, thereby overlooking their rich diversity of emotional expressions. In contrast, athletes from Ireland and New Zealand were shown smiling. Highlighting a bias that fails to accurately represent individual athletes and perpetuates simplistic and often misleading cultural stereotypes.
Such emotional biases are damaging, as they perpetuate one-dimensional views of entire populations. Like all individuals, athletes experience a wide range of emotions that we shouldn’t reduce to stereotypes.
Commenting on the study, Dr Kelly Choong, researcher and senior lecturer in advertising at Edith Cowan University, said: “The biases in AI are driven by human biases that inform the AI algorithm, in which AI takes it literally and cognitively. Human judgements and biases are drawn and presented as if factual in AI, and the lack of critical thinking and evaluation means the information is not questioned for its validity, just the objective of completing a task.
Biases will always lead to issues of equity, harmful generalisations, and discrimination. With society increasingly relying on technology for information and answers, these perceptions may end up creating real disadvantages for people of various identities. For example, the country’s association with certain sports may result in the perception that everyone in that country is prolific at it (Kenya – running; Argentina – football; Canada – ice hockey). These distorted ‘realities’ may also become embedded into individuals who believe these stereotypes and inadvertently reinforce them in real life.
Technology will find a way to better their algorithm and output. However, it will still be focused on completing a task, rather than offering a truthful representation. Society will need to question the validity and critically assess information generated by AI. Essentially, educating users will be paramount to the co-existence of AI and information, as well as the ability to challenge its output.”
Full results from the study can be found here
