The Gender Gap Also Extends to AI

The gender gap in the field of AI is very considerable, both in the workplace and in research. What is the origin of this inequality? What are the consequences? And how could it be reduced?

We have been talking about the gender gap in the technology sector for some time. And now, with the irruption of AI, we have discovered a new gap that should be corrected as soon as possible.

“The gender gap in AI largely reflects the inequalities in the technology sector in general that exist today. Traditionally, the field has been male-dominated, which can be seen in a lower presence of women in both STEM (science, technology, engineering and mathematics) education and AI-related careers. Although there are no up-to-date quantitative studies, it is easy to see that the gender gap persists, which suggests that the gap is going to be the same, or even greater, in the field of AI, due to the rapid evolution and specialization required in this area,” warns Marta Graño, professor of Leadership and Innovation at OBS Business School and author of the ‘Report on Gender Gap in the Age of Artificial Intelligence’.

For example, a few months ago we reported that women occupy a third (34%) of STEM positions in Spain and that men are 1.8 times more likely than women to work in occupations in these fields. Moreover, although more women are now graduating in STEM fields, not all of them end up working.

The gap in the field of AI is even wider than that already existing in the technology sector. Salomé Valero, head of Applications, Data and Artificial Intelligence at Kyndryl Spain and Portugal, points out some data that highlight the gender gap in AI. “According to UNESCO, only 30% of research academics in the field of AI are women. And this proportion drops drastically to 11% when it comes to managerial positions,” she denounces.

“These figures are also observed in the labor sector. The World Economic Forum notes that less than 25% of AI specialists are women, while in specific areas such as cloud computing and engineering, female representation is even lower, at only 14% and 20%, respectively,” he adds.

Cause of the AI gender gap

The causes behind this gender gap in the AI field are diverse and include socio-cultural, educational, and structural factors. “One cause may be the lack of representation of women in STEM disciplines, which has created a deficit in female talent in the AI field. According to Unesco, currently, only 22% of STEM students, less than 30% of STEM workers and less than 30% of scientific researchers in the world are women,” states Elena Viniegra, Cloud Director EMEA and Latam at NetApp.

“Similarly, there are entrenched gender biases in the hiring and promotion processes that can unintentionally favor men over women, thus perpetuating inequality in the workplace,” she notes.

The OBS Business School professor agrees with her, extending these biases beyond the workplace. Thus, she believes that “there are numerous obstacles that can hinder women’s access and progression in AI, from educational biases and biases in the admissions process to hiring and promotion practices in the workplace.”

In addition, Valero speaks of “the lack of female role models in the technology field, coupled with a predominantly male work culture and stereotypes about women’s capabilities in traditionally male-dominated areas.”

Graño also stresses this lack of role models. “The scarcity of visible and recognized women in the field of AI can diminish the aspiration and motivation of young women.” She also dwells on the barrier posed by gender stereotypes. “Traditional gender roles and expectations can discourage girls and young women from pursuing interests in technological and scientific fields.”

On the other hand, Viniegra remarks that “organizational cultures that are not inclusive of women can discourage their participation in the AI field, resulting in lower representation on development and leadership teams.”

Ultimately, we can conclude that the basis for this imbalance is cultural. “This gender gap in technology comes from the same place as in all other professions: because of culture and patriarchy. Culture has masculinized these jobs. For example, the developer’s job is a historically masculinized job. It is a job of focus and not focus. As a result, women are less attracted to it. This is not going to change by magic and, unfortunately, there are fewer and fewer women in these jobs. This line will continue if we don’t turn it around. Right now, for any job, we use technological tools, which makes technology more democratic, but technical positions are still male,” laments Fabiola Pérez, CEO and co-founder of MIOTI Tech & Business School.

Consequences of the gap

The first and most obvious consequence is the underrepresentation of women in AI research and development, as well as in jobs specialized in this technology.

Technology jobs are the most valued and the ones with the highest salaries. In other words, it’s another way of covering up a wage gap. If women are not in those professions – which are also exponentially growing professions – to position themselves in an interesting situation when it comes to growing in this industry, obviously, they are left behind in terms of salaries,” Pérez stresses.

But the existing gender gap in the field of AI not only poses a challenge from the point of view of labor equality, but can also have other very serious consequences.

“Beyond labor inequality, the gender gap in AI can have consequences such as biases in AI algorithms and systems and a limitation in innovation,” Graño warns.

If the historical data with which AI is trained is biased or contains stereotypes, these are going to be perpetuated,” he says. In addition, he stresses that “the lack of diversity in AI development teams can lead to the creation of systems with built-in biases, which has a negative impact on equity and social justice.”

Similarly, Valero emphasizes the need to form more diverse development teams, which “can better identify and address biases, creating more equitable and fair technologies for all,” thanks to their “different perspectives, experiences and approaches to the decision-making process.” “By addressing the gender gap, developers can work to identify and mitigate these biases, creating more unbiased algorithms that do not favor one gender over another,” he adds.

On the other hand, the OBS Business School professor recalls that “diversity fosters innovation”. “A lower participation of women can limit the variety of perspectives and creativity in the design and implementation of technological solutions,” she clarifies.

The perspective of half of the population is lost, and this is a problem,” adds the head of MIOTI. Likewise, Viniegra emphasizes the “underutilization of female talent”. “Women have unique skills and perspectives that they can bring to innovation and technological development. However, when they are denied access or discouraged from entering these areas, the opportunity to benefit from their creativity, analytical skills and problem-solving abilities is lost.”

What should we do?

The OBS Business School professor points out some measures that she believes could be effective in increasing the presence of women in the field of AI, as outlined in her report.

The first is inclusive education. “Promote STEM education from an early age, with an approach based on eliminating gender stereotypes.”

In addition, he insists on the importance of outreach and awareness. “Awareness and outreach campaigns should be used to highlight the achievements and contributions of women in AI. This helps to challenge stereotypes and attract more women to the discipline.”

She also recommends implementing mentoring and mentorship programs “that connect young women with successful AI professionals.” “Mentorships offer guidance, support and practical insight into the opportunities and challenges facing women in the industry,” she specifies.

She also believes it is critical to “disclose gender gap data.” “Faced with the differences that persist, we must show the relevant statistical and scientific information, show the reality, so that society is aware that there is still a long way to go. Because people can have different opinions, but we cannot have different data,” she says.

In addition, the head of NetApp believes that “there is a need for ethical governance of AI systems and tools.” “In addition to specific laws and policies, we need to train the people who develop and use these technologies, in addition to involving and empowering young people to become agents of gender equality,” she says.

Likewise, he points out that the prompts we give to AI systems should be “reviewed and adjusted to ensure that they reflect gender-equitable and unbiased perspectives.” “By providing female roles and guidance free of bias, we can get better quality results that are more aligned with an inclusive and diverse world,” he clarifies.

She also emphasizes that “it is necessary to have the financial support, political will and active participation of women in the spaces where AI tools are designed and regulated.” “This ensures that diverse perspectives and experiences are incorporated into the decision-making process, which contributes to a more ethical and equitable governance of AI,” she assesses.

On the other hand, the CEO and co-founder of MIOTI, thinks that this gap in AI could be mitigated by providing more applicability to the technology and making the tools themselves available to as many women as possible”-.

In this regard, she believes that generative AI could be the necessary catalyst. “From the point of view of using generative AI in a democratized way, we can see it as an enabler of changing the female presence within technology professions across industries. Before, maybe we were afraid to approach technology and that could be one of the reasons for the gender gap. But now, with generative AI, the reality is that technology is more approachable,” she explains.

“Generative AI becomes an enabler of creative thinking. Just as we’ve seen the gender gap in UX/UI specialists, graphic design or even video game design, which can be considered STEM studies, shorten a lot, with generative AI we’re going to see that improvement in the gender gap as well. Generative AI is still a technology that has recently been democratized with the advent of ChatGPT, so we’re all starting from scratch. And this can scale to all levels,” he adds.

What is being done?

Graño reviews that “in Spain there are initiatives aimed at encouraging the participation of women in technology and AI, such as MujeresTech and #SheMeansBusiness, programs and social networks that seek to inspire and connect women interested in technology; or WomanTechOver, an initiative that promotes the visibility and recognition of women in the technology sector.”

Valero talks about the initiatives being developed by Kyndryl in this direction. “Recently, we have developed robotics and programming initiation workshops aimed at girls and boys in Madrid and Barcelona, in collaboration with the United Way Foundation. These workshops are an integral part of our efforts to support and foster female talent in the field of technology,” she details.

In addition, she indicates that the company fosters “an inclusive and connected culture” and works to build “diverse and representative teams at all levels and geographic regions” of the company. It also has Kyndryl Inclusion Networks, groups that provide a supportive environment for its members and contribute to the advancement of their respective communities.

NetApp also has initiatives to promote technology education and data literacy as tools to address the gender gap, such as its ‘Data Explorers’ program. “It offers students the opportunity to delve into education-related data and reflect on their own context in the classroom. Through this process, they can develop conclusions about what they would like to change in education, which empowers them to take action and contribute to improving their educational environment.”

He further explains that the company takes responsibility for promoting greater demographic, physical and cognitive diversity in its work environment, “creating opportunities that foster diversity, equity, inclusion and belonging on a daily basis.” “Our culture of inclusion is a palpable reality, which promotes an unbeatable work environment,” he says.

For her part, Perez notes that MIOTI works with different women in technology associations “to highlight the opportunity that AI, and especially generative AI, brings to all types of departments and groups.”

“We have a responsibility to directly encourage women to study technology careers and encouraging girls around the age of 14 to see technology as attractive is fundamental. That is why we are organizing conferences, workshops, and meetings with girls and women technologists and entrepreneurs on our campus,” she highlights.