forbes.com
AI Facial Recognition Discriminates Against Millions With Facial Differences
Face Equality International's 'Visibly Invisible' campaign highlights how AI facial recognition technology discriminates against 72 million people globally with facial differences, causing issues with social media, banking, and travel.
- What immediate impact does the failure of AI facial recognition technology have on individuals with facial differences?
- A recent survey by Face Equality International reveals that 32% of individuals with facial differences face issues uploading photos to social media, 21% experience problems with banking apps, and 12% cannot use e-passport gates. These difficulties highlight the discriminatory impact of AI-based facial recognition technology.
- How do the findings of Face Equality International's survey illuminate the systemic issues surrounding the use of AI-powered facial recognition?
- The 'Visibly Invisible' campaign, launched by Face Equality International, aims to address the widespread challenges faced by people with facial differences due to AI facial recognition systems. The campaign uses survey data showing significant issues with social media, banking apps, and airport security, emphasizing the need for improved technology and legislation.
- What legislative and technological changes are necessary to mitigate the discriminatory effects of AI facial recognition on individuals with facial differences?
- Future implications of this technology's expanding use in gaming, retail, and employment could further marginalize individuals with facial differences unless significant improvements are made. Legislative action is crucial, as current disability frameworks often fail to protect those with facial differences who lack functional impairments but still experience significant discrimination.
Cognitive Concepts
Framing Bias
The article frames the issue primarily from the perspective of individuals with facial differences experiencing discrimination. While this is important, the framing could be improved by including more balanced perspectives and acknowledging the complexities of the technology and its development. The headline and introduction strongly emphasize the negative impact, setting a tone that prioritizes this aspect of the story.
Language Bias
The language used is generally neutral, although terms like "non-standard faces" and "overt marginalization" might carry slightly negative connotations. More neutral alternatives could include "different facial features" and "significant discrimination.
Bias by Omission
The article focuses heavily on the negative experiences of individuals with facial differences using facial recognition technology, but it omits discussion of potential benefits or alternative technologies. It also doesn't explore the perspectives of developers or companies creating this technology, limiting a comprehensive understanding of the issue. While acknowledging the small sample size of the survey, it doesn't discuss the limitations this might place on generalizing the findings to the wider population of individuals with facial differences.
False Dichotomy
The article doesn't present a false dichotomy, but it could benefit from acknowledging that the challenges faced by individuals with facial differences don't negate the potential benefits of facial recognition technology for others. A more nuanced approach would discuss both sides of the issue more evenly.
Sustainable Development Goals
The article highlights how AI facial recognition systems discriminate against individuals with facial differences, exacerbating existing inequalities and marginalization. This technology, despite offering convenience to many, excludes a significant population due to its bias in recognizing "normal" facial features. The resulting exclusion from everyday services like banking apps, social media, and airport security creates significant inconvenience and emotional distress, furthering societal inequalities.