Artificial Intelligence (AI) development is inadvertently incorporated with prejudices due to the lack of diversity in the tech industry

Artificial Intelligence (AI) development is inadvertently incorporated with prejudices due to the lack of diversity in the tech industry

In the rapidly growing field of Artificial Intelligence (AI), diversity and inclusion (D&I) remain a significant concern. According to recent data, underrepresented groups, such as Blacks, Hispanics, and Native Americans, hold a disproportionately small percentage of AI technical jobs compared to their representation in the US workforce as a whole.

For instance, Hispanics hold approximately 9% of AI technical jobs, while they make up over 18% of the US workforce. Similarly, Black workers account for about 8% of the technical AI jobs, compared with nearly 12% of the overall US workforce.

This disparity is also reflected in the education sector, with about 60% of public high schools offering AI classes. However, underrepresented groups are less likely to have access to these opportunities.

Efforts to address this imbalance and make AI more representative of American society are facing resistance, particularly from the Trump administration. Critics argue that government criticism of tech and social media companies as being too available to civil rights and human rights messages could hinder progress in creating opportunities in AI for marginalized groups.

Professor Safiya Noble, of the University of California Los Angeles, shares this concern. She worries that the government's attack on D&I programs may undermine efforts to create opportunities in AI for underrepresented groups.

Despite these challenges, the importance of D&I in AI cannot be overstated. A diverse and inclusive workforce enables the creation of AI systems that are more fair, less biased, and more applicable to a mass and varied population. This is achieved by incorporating multiple perspectives throughout the design and decision-making process.

Diverse teams are better equipped to identify biased data and discriminatory outcomes, reducing the risk of AI reinforcing narrow or harmful stereotypes. Inclusion fosters fairness by addressing discriminatory outputs that arise when AI systems are built by homogenous teams whose experiences may not represent broader society.

Moreover, diverse teams bring a greater variety of perspectives, which enhances creativity, innovation, and problem-solving, leading to AI solutions that are relevant to different demographics and use cases. Workforce diversity in AI can also improve employee engagement and reduce turnover, creating a more positive environment for sustained focus on fairness and inclusion in AI development.

Companies that actively pursue diversity and conduct regular audits, maintain human oversight, and develop transparent AI are better positioned to detect and correct biases, leading to more equitable AI-driven outcomes.

Unfortunately, progress towards gender equality in the AI workforce is slow. From 2021 to 2024, the number of women working in AI globally increased by only 4%. Women currently represent 26% of the AI workforce, according to a UNESCO report.

The underrepresentation of women, along with racial and ethnic minorities, in the AI workforce is a global issue. The Thomson Reuters Foundation, the charitable arm of Thomson Reuters, has highlighted this concern.

One promising figure is Maya De Los Santos, an Afro-Latina woman with degrees in computer and electrical engineering. She hopes to forge a career in AI, demonstrating the potential for a more diverse and inclusive AI workforce.

However, the US government's termination of D&I offices and ban on federal contractors from using affirmative action in hiring could hinder efforts to increase diversity in the AI workforce. This is a concern for experts and observers, who argue that such moves could undermine efforts to create opportunities in AI for underrepresented groups.

In conclusion, promoting D&I in the AI workforce is crucial for the development of fair, unbiased, and relevant AI solutions. A diverse and inclusive AI workforce not only reduces the risk of AI reinforcing harmful stereotypes but also leads to AI products that serve all users more effectively and help reduce inequalities that AI might otherwise amplify.

1. The disparity in representation of underrepresented groups, such as Blacks, Hispanics, and Native Americans, in AI technical jobs can be seen in contrast to their proportion in the US workforce as a whole.
2. Companies that prioritize diversity in AI workforce by conducting regular audits, maintaining human oversight, and developing transparent AI systems are better positioned to detect and correct biases.
3. Progress towards improving gender equality in the AI workforce is slow, with women's representation currently at 26% globally, and the US government's termination of D&I offices and ban on federal contractors from using affirmative action in hiring could hinder efforts to increase diversity.
4. Diverse teams in AI workforce can bring a greater variety of perspectives, which enhances creativity, innovation, and problem-solving, leading to AI solutions that are relevant to different demographics and use cases.

Read also: