In the digital age, artificial intelligence has slowly become everyone’s guide to success, where we stop using our own intelligence and borrow that of these man-made systems to navigate our lives. Worried about an upcoming interview? Ask the AI. Want to try a new recipe? Ask the AI. Are you planning to send a love message to your partner? Ask the AI. But who manages the AI? We must lift the veil on a truth that has been demonstrated beyond doubt: there is a lack of diversity in these technological advances. From biased hiring algorithms to discriminatory healthcare practices, these teams are building systems that have a natural bias against women and people with darker skin.
Unbeknownst to many, artificial intelligence is just as prone to racial and gender bias as society. Due to existing gender and racial biases in communities, the same challenges seep into the training data used in AI generation systems that yield biased results, often unfavorable to marginalized communities. A study by Joy Buolamwini and Timnit Gebru found that commercial facial recognition systems from big tech companies like Amazon produced erroneous results for women and people with darker skin. Research also found that voice recognition systems have higher error rates when women are involved than men. This will inevitably lead to frustration and exclusion of these people.
Take, for example, the Amazon AI Recruiting tool. The tool was trained on CVs sent to the company over 10 years. Since tech is a male-dominated industry, the tool was trained on this data, and it was inherently biased and unintentionally chose male candidates over female candidates. Amazon later made changes to account for these errors, but ultimately lost hope of making the tool gender neutral. This is a crucial example of how women and some minorities are being excluded from the future field of AI.
AI systems learn from the data they receive, meaning that if the data used in a facial recognition system is comprised primarily of white men, it will generate inaccurate results when dark-skinned women are involved. A UNESCO study also found that broad linguistic patterns associate women using keywords such as “home”, “children”, while men are associated with “career” and “executive”. This gender bias comes from the fact that the system has been trained to believe that women are supposed to stay at home and take care of children while men work in executive offices.
When LLMs were asked to generate a narrative of people of different ethnicities, racial biases emerged. British men were given occupations such as “driver”, “doctor” and “bank clerk”, while Zulu women were assigned roles such as “maidservant” and “housekeeper”. In 2021, UNESCO proposed a global normative framework called Recommendation on AI Ethics which calls for specific measures including investment in targeted programs in marginalized communities to increase opportunities for girls and women in STEM and ICT fields. Large technology companies like IBM have endorsed this and are working to improve their AI tools accordingly. UNESCO also launched a program called Women4EthicalAIwhich is a collaborative platform to help governments and businesses fulfill their mission to ensure that women are not excluded from the development and deployment of AI systems.
Facial recognition technology, which has been hailed for improving security, has also been pointed to some inherent biases. Joy Buolamwini, founder of the Algorithmic Justice League, began researching this area due to some of her own personal experiences. While studying at MIT, she discovered that some facial recognition technologies couldn’t identify her until she put on a white mask. Additionally, some systems at these large companies were unable to recognize Oprah, Serena Williams and Michelle Obama as female subjects and identified them as male. This sparked a broader discussion about the biases of AI systems sold to governments and law enforcement.
The consequences of AI bias go far beyond the downsides: they can often exacerbate systemic inequalities, further marginalizing already vulnerable populations. For example, a biased facial recognition system may falsely identify a suspect, which can perpetuate a wrongful conviction. Police could end up profiling the wrong person because AI tools couldn’t accurately identify the correct all-black man as a suspect. In healthcare, AI tools can cause misdiagnosis of patients from minority communities, thereby exacerbating racial bias in society.
The best way to solve this problem is to recognize that this bias exists. Once these organizations are ready to bridge the gap, they must prioritize diversity and inclusion. When designing and training algorithms, teams should ensure they include diverse perspectives from different backgrounds and expertise. Algorithms should also be tested frequently and thoroughly to eliminate bias and audited to identify discriminatory patterns. A broader and more diverse perspective ensures deeper analysis. Additionally, all needs of all stakeholders will be met without leaving anyone behind. When teams are less diverse, they are more likely to have blind spots and unconscious biases that can negatively influence the design of AI systems. Therefore, it is more helpful if the team is fully diverse and inclusive.
The best way to solve this problem is to recognize that this bias exists.
The journey to eliminating bias in AI is fraught with challenges, but it is a mandatory journey that will help develop a more equitable society. By identifying, confronting and correcting these biases, we can pave the way for new AI systems that serve the greater good of all, without spreading discrimination and racial bias. Through an unwavering dedication to conscious inclusion, we can usher in a new era of gender- and race-friendly artificial intelligence, where no one is left behind.
We should also aim to balance gender equality in the workforce so that the data available for training AI models is more inclusive of the entire population. This is not an easy task, but we must embark on this journey so that the technology industry offers products that equitably benefit all members of society.
