As AI transforms hiring, promotions, and daily work, trusting it to be objective is a dangerous gamble. The work of trailblazers like Joy Buolamwini, Timnit Gebru, and Rumman Chowdhury (Buolamwini, 2019) shows a system flawed at its core—a system that, if left unchecked, could systematically undermine the careers of women of color and first-gen professionals.
The “Coded Gaze”: When Technology Doesn’t See You
Buolamwini’s (2019) research at MIT uncovered the “coded gaze.” She found that facial recognition software from giants like IBM and Microsoft failed to detect her dark-skinned face unless she wore a white mask. This wasn’t an anomaly; it was a symptom. These systems were trained overwhelmingly on images of light-skinned men.
The consequences are stark:
- Error rates for darker-skinned women soared to over 35%.
- Systems consistently misgendered Black women, while accurately identifying white men.
This technical failure has human costs. It creates an “exclusion overhead”—the extra burden marginalized people bear to be seen and recognized by technology (Buolamwini, 2019).
From Faces to Futures: How Bias Scales
This bias isn’t confined to facial recognition. Gebru’s research into large language models (the foundation of tools like ChatGPT) found they perpetuate and amplify societal stereotypes.
- Prompts like “the woman worked as” generated derogatory associations, but “the man worked as” did not.
- These biased models are then applied to predictive analytics for hiring, loan applications, and even criminal sentencing.
The lack of diversity in the teams building AI and the data used to train it has created technology optimized for a narrow slice of humanity. As Gebru found, when you are not at the table, you are on the menu.
Your Role as a First-Gen Leader: From Awareness to Action
This is not a remote, technical problem. It’s a workplace reality. So, what can we do?
- Become AI-Literate, Not Just a User: Understand that AI outputs are suggestions, not truths. Question the data and the potential biases behind the tools your company uses.
- Advocate for Transparency: Support the work of leaders like Chowdhury, who push for systems to be audited and opened for analysis. In your organization, ask questions about how AI tools are vetted for bias.
- Leverage Your Unique Perspective: Your experience as a navigator of complex systems gives you a critical eye for when something is unfair or “off.” This is a form of judgment AI lacks. Use your voice to challenge biased outcomes.
The warnings from these women are not abstract. They are a call to action for every professional who cares about equity. By understanding the landscape, we can switch from being passive subjects of technology to active shapers of a more ethical future.
Sources
Buolamwini, j. (2019, February 7). Artificial intelligence has a problem with gender and racial bias. Here’s how to solve it. TIME. https://time.com/5520558/artificial-intelligence-racial-gender-bias
