AI Reinforces Unrealistic Body Ideals for Athletes, Study Finds

A recent study conducted by researchers at the University of Toronto revealed that artificial intelligence (AI) is perpetuating unrealistic body ideals, particularly for athletes. The study analyzed AI-generated images of both male and female athletes and non-athletes, uncovering a trend of exaggerated and likely unattainable body standards.

Athletes have long faced societal pressures to conform to specific body images, often characterized as thin, muscular, and conventionally attractive. These pressures come from coaches, competitors, spectators, and the media, emphasizing appearance over athletic performance. The consequences of such objectification include negative body image, poor mental health, and decreased performance in sports. With the increasing prevalence of AI on social media, understanding how these technologies depict bodies has become essential.

AI Depictions: A Narrow View of Athleticism

The research team generated 300 images using various AI platforms, including DALL-E, MidJourney, and Stable Diffusion, to examine how athlete and non-athlete bodies are presented. The study documented demographic factors, body fat levels, muscularity, clothing fit, and facial attractiveness.

The findings revealed that AI-generated images of male athletes predominantly featured young, lean, and muscular individuals, with 93.3 percent appearing young, 68.4 percent lean, and 54.2 percent muscular. For female athletes, the images presented a similar trend, with 100 percent depicted as young, 87.5 percent thin, and almost all (87.5 percent) wearing revealing clothing. Notably, images of athletes were almost universally lean (98.4 percent) and muscular (93.4 percent), dressed in tight and revealing exercise gear.

In contrast, images of non-athletes showcased a broader diversity of body sizes and clothing types, often featuring looser-fitting clothing. Alarmingly, when requesting images solely of “athletes,” researchers found that 90 percent of the generated images were male. Furthermore, none of the images included individuals with visible disabilities or larger body types.

Consequences of Distorted Body Ideals

The implications of these findings are significant, especially given that over 4.6 billion people use social media, with 71 percent of the images generated by AI. This exposure can foster self-objectification and the internalization of distorted body ideals, leading individuals to engage in unhealthy behaviors, such as excessive dieting or over-exercising, in an attempt to conform to these unrealistic standards.

Research indicates that negative body image affects not only personal well-being but also academic and athletic performance. While maintaining an active lifestyle can enhance body image, the reverse is often true; a poor body image can exacerbate dropout rates in sports and physical activities.

The absence of diverse representations, particularly regarding individuals with disabilities, is especially concerning. Approximately 27 percent of Canadians over the age of 15 live with at least one disability, making the lack of representation in AI-generated imagery a glaring oversight.

The study highlights that AI draws from a biased database of existing media, recycling and reinforcing harmful appearance ideals rather than offering diverse and realistic representations of human bodies. This reliance on narrow standards leads to a lack of diversity in the portrayal of bodies, contributing to a cycle of unrealistic expectations and decreased body satisfaction.

As creators of the visual content that informs AI systems, society bears responsibility for ensuring that these technologies do not perpetuate forms of discrimination, including ableism, racism, fatphobia, and ageism. Users of AI-generated imagery should take care in crafting prompts and critically assessing the images they encounter.

To combat the proliferation of narrow body standards, it is crucial for society to advocate for a more inclusive representation of all body types in AI-generated images. By insisting on diverse portrayals, there is potential to reflect reality more accurately and foster a healthier public perception of body image.

In closing, as AI-generated images continue to shape our media landscape, it is essential to remain aware of the narratives these images promote and to challenge the limited views of athleticism they often convey.