Survey: Women are eager to use gen AI tools, even as they are ‘vastly underrepresented’ in AI

Women have long been underrepresented in the tech field — and AI is no different. But that’s in no way a reflection of their interest in technology and eagerness to adopt generative AI tools. 

In fact, 68% of female respondents to a new survey from Women Go Tech have used at least one AI tool — hands down, the No. 1 preferred is ChatGPT — and 61% said they wanted to learn more about AI tools and applications.

Still, women can be hesitant initially due to insecurities fueled by age-old biases and discrimination, as well as fears about data privacy and ethical use of the technology, the survey discovered.  

“Women are generally very interested in AI and its potential applications but often hesitate to engage deeply with AI tools,” Women Go Tech writes in a report summarizing the findings. “The technical nature of AI and prevailing gender stereotypes may lead many women to feel less confident about their ability to effectively use and understand AI tools. This lack of confidence is a significant barrier that prevents them from fully engaging with the technology.”

Discrimination, bias, ‘Matilda Syndrome’ still persist

Supported by Google.org and the Organization for Security and Co-operation in Europe (OSCE), the survey polled 5,400 respondents from 13 European countries. Women were split into four categories: those interested in tech careers (31%), those not interested in tech careers (30%), those who have worked in tech for more than 2 years (27%) and those who have worked in the tech sector for less than two years (12%). 

Of the aspiring technologists, 77% were interested in AI, while 64.6% of tech sector novices and 63.9% of experienced talent said they wanted to learn about the technology. 

Beyond ChatGPT, those who had tried out the technology tinkered with OpenAI Codex, Bard/Gemini and GPT-4. The most popular use cases were translation, navigation/travel and search for answers. 

However, one-quarter of respondents — across both new and experienced talent groups — reported that they did not feel entirely comfortable with their technical skills. 

“It would be hard to believe that specialists would find their general tech skills lacking, so it is important to consider how factors like low confidence, lack of encouragement or access to education might also influence someone’s self-esteem,” the report emphasizes. 

Women often underestimate their abilities due to “pervasive stereotypes and societal expectations” that can deter them from technology, which in turn leads to a lack of female representation and ongoing bias, the report notes. Also, many women experience “imposter syndrome,” where they do not feel like they belong in their roles despite having the necessary skills and qualifications. 

“We have amazing female engineers, mathematicians,” said Ana Prică-Cruceanu, chief SDG strategy officer -AFIST, Women4Ethical AI at UNESCO. “We need to highlight these role models. We need to talk about the Matilda effect [when the work of female scientists is attributed to men]. We need to make sure they understand that you can start in tech at any age.”

Indeed, 28% of respondents reported that they had experienced discrimination, and 32% were concerned that they might experience it in the future.

Paradoxically, though, 46.5% of respondents “strongly agreed” that AI technologies are designed with diverse user perspectives in mind. This is quite the divergence from experts who say that women are “markedly underrepresented” in AI development and training datasets, which can lead to discrimination and AI bias. 

This can lead to a lack of role models, mentorship opportunities and a “cycle of continuous exclusion,” the report notes. 

“I think it’s important that women support each other in the field,” said Lili Kuusk, CEO of the Estonian Association of Information Technology. “They may be alone in one company, but overall, in the field, they are not alone.”

Vague company policies

Many organizations are rushing to implement AI tools — but have yet to establish clear policies around the technology. Reflecting this, 30% of survey respondents said they were not sure about their employer’s policy and just 8% said they have received guidance on how to use AI at work. 

Conversely — and not surprisingly — employer policies that allow and guide AI use correlate to the highest usage rates (69.4%), followed by clear, prescriptive policies (64.1%).

Other factors hindering survey respondents from using AI were a lack of clear understanding about how the technology works and concerns about data privacy and ethical use. 

“Companies should ensure all employees are aware of AI policies and how these tools can facilitate their work,” the report authors emphasize. 

This includes providing easy-to-understand learning materials, establishing support networks for sharing AI experiences and educating on AI’s privacy and safety aspects. It is also important to highlight how AI can enhance rather than replace human interactions, the report authors note. 

To support those highly interested in using the technology, employers might emphasize AI as a “career enhancer” and provide tailored learning paths with hands-on projects or certifications that align with their career aspirations, the authors suggest. 

Meanwhile, employers might help more advanced “power users” to become mentors who lead workshops or present at conferences. They might also consider offering specialized courses in advanced AI fields such as machine learning (ML) and robotics, and encourage participation in hackathons and collaborative projects that provide practical expertise and allow them to showcase their skills. 

The authors conclude: “By providing clear frameworks and support, employers can effectively encourage women to engage more deeply with AI technologies, enhancing their skills and confidence in utilizing these tools in their professional roles.”