AI for All: The Case for More Women Leaders in AI

Who’s behind your favorite AI tools? In most cases, men.

The AI transformation is shifting how we work and live, with no one knowing the full extent to which it will change our lives. AI won’t just change how we approach problem-solving—it may well change how we think. Simultaneously, leadership gender parity remains years away, with women holding less than 10% of CEO roles in the S&P 100 and the FTSE 100.

To better understand how the leadership gender diversity gap is playing out within this dynamic space, Russell Reynolds Associates took a deep dive into 39 AI application and platform organizations, analyzing the gender diversity of their senior leadership teams (see below for complete methodology). Additionally, we analyzed data from our Global Leadership Monitor, which captures the perspective of leaders across industries, including how they are interpreting AI's potential impacts on their organizations.

Our analysis found that women leaders’ representation is equally poor within these AI-focused organizations (Figures 1 and 2), with women holding only 30% of the overall leadership roles and 10% of the CEO and top tech toles at these organizations.

Figure 1: Women leaders’ representation within AI platform and application organizations

^{Source: RRA Proprietary Analysis, select AI platform & application organization leadership teams, 2024 (n=39 companies, 426 executives); data on executive teams was collected via company websites and LinkedIn and was categorized and analyzed in Q4 of 2024.}

The majority of women leaders in these organizations are clustered in HR and legal, holding 86% and 55% of these roles, respectively (Figure 2). These functional roles rarely hold the same level of influence as traditional “CEO feeder” roles, nor do they typically lead to the top job.

Figure 2 : Gender diversity in AI platform/application organization executive roles

^{Source: RRA Proprietary Analysis, select AI platform & application organization leadership teams, 2024 (n=39 companies, 426 executives); data on executive teams was collected via company websites and LinkedIn and was categorized and analyzed in Q4 of 2024. Only included roles in which sample sizes were >20. Given the range of organizations analyzed, SLT size and composition varied significantly.}

Surprising? Not particularly. Concerning? Absolutely.

AI’s proliferation has resulted in a complex, rapidly-changing vendor landscape that is already having massive implications for our workforce. With how quickly things are changing, we know many leaders are making decisions about this technology without fully understanding it. And leaders are even less optimistic about their organization’s accompanying direction around AI, as our Monitor found that only 27% agree that their organization has provided the right level of guidance to harness generative AI ethically and safely, and only 24% agree that they have the right processes in place to protect against AI misuse. This means that organizations are relying on AI vendors to design useful and ethical technologies with appropriate safety controls and bias mitigation and detection protocols.

The problems AI organizations face with bias in their tools is well documented. There are numerous instances of AI producing results that perpetuate stereotypes and bias in wide range of use cases, including healthcare, applicant tracking systems, online advertising, image generation, and predictive policing tools. Despite AI appearing neutral, it's made by humans. This means it internalizes and iterates on all the same biases we have - including gender bias.

But what we are talking about goes beyond distasteful or comically incorrect outputs. AI tools and their forebearers are applied to decisions that have real world consequences for individuals and communities – be that decisions on job applications, mortgage and loan applications, or healthcare. As such, the leaders of AI organizations wield a huge level of influence. With that comes great responsibility.

Large Language Models (LLMs) are trained on vast existing data sets designed to represent natural language and function. But historically, data collection has centered men’s experience and overlooked women. This male centricity is rarely deliberate—it’s typically the function of cost, convenience, or a simple lack of consideration. But whether intentional or not, biased product design (stemming from incomplete or inaccurate data sets) has resulted in serious disadvantages for women and their wellbeing, impacting:

^{For more on the adverse effects on women caused by the gender bias in big data collection, see Caroline Criado Perez’ Invisible Women: Data Bias in a World Designed for Men, New York, NY: Abrams Press, 2019.}

AI can help us guard against repeating these issues – but only if designed with the appropriate guardrails, and with input from everyone it aims to reflect and protect. While improved gender balance in leadership teams is by no means a silver bullet, an organization’s leadership helps shape those safeguards, answer big questions around where their company should focus, and outline what problems they’re aiming to solve with this technology.

Low representation in technology functions leads to few women in top roles at AI organizations

While every senior leadership role is important to a company’s healthy functioning, if AI is to truly meet the future needs of society, more balanced gender representation in its technical design and governance will be critical to stopping the perpetuation of unintended and embedded gender biases.

But currently, women leaders are severely underrepresented in technology leadership roles in AI companies, holding a mere 22% of the product, engineering, and science roles observed (Figures 2 and 3). And we observed only four women CEOs and four women CTOs across these 39 organizations.

Figure 3: Women leaders’ representation in select key roles at AI organization

^{Source: RRA Proprietary Analysis, select AI platform & application organization leadership teams, 2024 (n=39 companies, 426 executives); data on executive teams was collected via company websites and LinkedIn and was categorized and analyzed in Q4 of 2024.}

While the small sample size makes broad conclusions difficult, our findings suggest that when women helm AI companies and/or their tech teams, they’re more likely to achieve gender-balanced leadership. Of the four women-led AI organizations we examined, two have achieved gender parity and three are more gender diverse than the average AI leadership team (meaning their senior leadership teams are comprised of over 30% women.)

Conspicuous in their absence: Only two AI organizations have achieved gender parity on their leadership team

While it’s important to understand gender representation across AI’s overall executive population, it’s perhaps even more crucial to understand how it varies between organizations. With that, we also analyzed the leadership teams of these 39 AI platform and application organizations.

As of Q4 2024, four of these organizations’ senior leadership teams were comprised of all men, and seven have only one woman executive onboard (Figure 4). Nineteen of these C-suites have less than 25% women, and 30 are comprised of less than one-third women.

Only two of the 39 organizations analyzed have senior leadership teams comprised of over 50% women. As noted above, both of these organizations are led by women.

Figure 4: AI platform/application leadership team gender diversity snapshot

^{Source: RRA Proprietary Analysis, select AI platform & application organization leadership teams, 2024 (n=39 companies, 411 executives – noting execs listed on organization’s website ONLY); data on executive teams was collected via company websites and LinkedIn and was categorized and analyzed in Q4 of 2024.}

Less represented, but more concerned: Women leaders are generally more worried about AI’s societal risks than their male counterparts

According to RRA’s Global Leadership Monitor, which gathers perspective from leaders across industries, women leaders at every level (next generation, C-suite, CEO, and board directors) are more likely to be concerned about AI’s most pressing issues in the workplace, including misinformation perpetuation and introducing bias into talent processes (Figure 4).

Figure 4: Women leaders across levels are more likely to express concern about AI’s societal impacts

^{Source: RRA H1 2024 Global Leadership Monitor, n=1,432 board, C-level, CEO, and next gen leaders, 2024.}

^{Source: RRA H1 2024 Global Leadership Monitor, n=124 C-suite women, 315 C-suite men}

It stands to reason that women leaders – who’ve likely had to contend with more gender bias throughout their careers than men – would be more attuned to AI’s bias risks, and why they’re more likely to be in favor of stronger regulation to ensure its safe use.

We’ve already seen many generative AI product launches miss the mark with women and underrepresented minority groups, resulting in embarrassing claw backs for their organizations. Given what we know about women leaders’ poor representation in AI organizations—particularly in roles that influence product design—one has to wonder if these missteps could have been avoided if more individuals who were keener on asking tough questions about bias and misinformation mitigations were in the room during the tool’s creation. Of course, this is not simply a question of representation, but also one of capability and organizational culture.

What’s next in the journey towards AI for all?

While bias may be an inescapable fact of life, it does not need to be an unavoidable aspect of new technologies. In fact, when designed thoughtfully, these technologies can help us solve these issues. While there are many steps AI organizations can take to ensure this outcome, ensuring that their leaders represent everyone for whom they’re developing this tech is key.

No company can outperform its leadership. And research has long shown that gender-balanced leadership teams produce better outcomes. Findings from our RRA Artemis movement indicate that the holistic way in which women leaders approach problem-solving and leadership is even more important in the increasingly volatile landscape of work. For technology organizations aiming to tap into the power of AI for all via their leadership teams, consider the following.

• Start succession planning early: Our research shows that, despite being equally effective, women are less likely to believe they’re qualified for leadership roles, as they tend to underestimate their skills. Counteract this and AI’s leaky tech leadership pipeline by inviting women into the process early and often. Not only will this help broaden the pool of next generation women leaders, when all qualified applicants know of and are in the running for leadership positions, it sends the message that everyone is capable and under consideration.
  
  
• Equip senior leaders to address common biases in the talent identification and succession process to transform CEO succession pipelines: Review the language around the objectives and competencies, bearing in mind that women are primarily motivated by impact (compared to status, pay, etc.) Related, ensure assessment teams truly understand—and value—men and women’s different experiences and motivations at work.
  
  
• Leverage AI itself to accelerate equitable outcomes: AI organizational leaders can embed equity steps into the talent sourcing process by leveraging AI to identify a wide range of participants, efficiently scan the relevant employee population, systematically assess applicants by matching assessment data with future potential roles, and tabulate individual calibrations to provide a selection recommendation that is not influenced by group-think. Additionally, leaders can leverage AI tools to quickly understand if different populations are being hired, retained, and promoted at comparable levels, thus empowering them to make leadership succession pipelines more equitable.
  
  
• Consider external development programs and joining communities of like-minded leaders: Sometimes, internal mentorship and advocacy is not enough to expose next-generation leaders to the range of experience required to progress to the next level. When this happens, consider investing in external development programs for leaders who are historically underrepresented in your industry. Additionally, joining a community of like-minded tech leaders offers opportunities for all leaders to discuss inclusive strategies and better outcomes. This allows leaders to network with their peers, gain new understanding that benefits both them and the business, and shows top performers that their organization is willing to invest in them for the long haul.

Additionally, we analyzed data from RRA’s H1 2024 Global Leadership Monitor, which is an online survey of executives and non-executives that gathers the perspective of leaders on the impact of external trends on organizational health and their leadership implications (first launched in 2021). Russell Reynolds Associates surveyed its global network of executives using an online/mobile survey from 4 March to April 1 2024. Data from previous Global Leadership Monitor surveys were deployed in February/March 2021, March 2022, October 2022, March 2023, and September/October 2023.

Authors

Leah Christianson and Tom Handcock of RRA’s Center for Leadership Insight conducted the research and authored this report.

LEARN MORE ABOUT THE AUTHORS AND THE CENTER FOR LEADERSHIP INSIGHT

Acknowledgements

The authors would like to thank the following contributors from Russell Reynolds Associates’ Technology and Board & CEO Advisory practices for their time and valuable perspectives:

Fawad Bajwa leads Russell Reynolds Associates’ AI, Analytics & Data Practice globally. He is based in Toronto and New York.

David Finke is a senior member of the firm’s Technology and Board & CEO Advisory practices. He is based in Palo Alto.

Margot McShane co-leads Russell Reynolds Associates’ Board & CEO Advisory practice in the Americas, and is the co-founder of RRA Artemis. She is based in San Francisco.

Hetty Pye is a senior member of Russell Reynolds Associates’ Board & CEO Advisory practice, and is the co-founder of RRA Artemis. She is based in London.

Tuck Rickards is a senior member and former leader of Russell Reynolds Associates’ Technology practice. He is based in San Francisco.