Women Go Tech, the NGO facilitating gender equality and women’s participation in tech through mentorship and career support, has published Women and AI: challenges, perceptions, and perspectives in the CEE region. Spearheaded by Women Go Tech and supported by Google.org and the OSCE, the large-scale study surveyed more than 5,400 women across 13 European countries. It is the first study of its kind in the CEE region to survey real respondents, rather than relying on conclusions and insights based on expert opinions.
AI Should Get to Know Women Better
Women are significantly underrepresented in AI worldwide, both in the context of professional roles and as subjects within data sets essential to training AI systems. This pervasive underrepresentation on the professional side hinders opportunities for mentorship and perpetuates a cycle of continuous exclusion. Within AI-training datasets, the underrepresentation of women creates discriminatory biases that can impact matters as serious as healthcare diagnostics, which can fail to account for the specific needs and characteristics of the female body. In other words, AI cannot be ethical and ethically used without women’s participation, but that is already lagging behind.
To provide a holistic picture, researchers surveyed four categories of individual: women who have worked in the tech sector for more than two years (27%), women who have worked in the tech sector for less than two years (12%), women who work in other sectors but are considering a tech career (31%), and women who do not work in tech nor would consider working in tech (30%).
68% of Respondents Use AI Tools, 61% Are Interested to Do So
Throughout the data collection process, the study repeatedly disproved the stereotype that women are less interested in technology or AI as their male counterparts. 68% of respondents had used at least one AI tool previously, and 61% of respondents expressed interest in learning more AI tools and applications. Bridging the gender divide wasn’t at the forefront of their minds or motivations – in fact, the women surveyed were less likely than AI experts to perceive AI technologies as biased.
Instead, these women tended to have different negative perceptions and false beliefs about AI, often related to job replacement, data privacy, and the potential impact of AI on human interactions. Other barriers were simply practical: 12% of the 61% of women who expressed interest in further AI-related training said they lacked the availability and time. Even half of the respondents who said they were not interested in further education mentioned time constraints as a barrier.
Women Want to Learn AI, Where Should They Start?
In total, 68% of respondents are interested in studying AI or working in an AI-related field. However, only half of them intend to do so. This reveals the study’s biggest takeaway: women absolutely want to devote time and energy to learning AI, however a one-size-fits-all approach will prove insufficient for integration. And with only 8% of companies maintaining clear AI guidelines and guidance, individual women often feel unsupported and at a disadvantage when working with AI. This is bad for integration and for productivity, as companies only stand to benefit from the efficiency and seamlessness skillful AI use provides.
The study divided its respondents into four categories with distinct sets of needs for integration. The first are the unengaged (38%), who need basic educational workshops and demonstrations of AI’s practical problem solving on an everyday level. The second are reluctant beginners (9%), who should be encouraged through access to relevant AI applications on familiar platforms and tools. The third are tech wannabes (21%), or basic AI users who can be made into advanced users through high employer engagement and training, certification programs, tailored learning paths, hands-on projects, and mentorship. And the fourth are power users (31%), who need an enriching and supportive environment to transition from user to developer via hackathons, specialized courses in advanced fields, and collaborative projects.
According to the study, women’s general disinterest in AI stemmed primarily from lack of confidence, whereas a lack of understanding about how the technology functions is still one of the key barriers to women’s engagement. Conversations and education about AI must differ based on the women in question. The future of AI will only be equal if its present challenges are properly understood.
