It’s a Power one.
A line from the G7 Leaders’ Statement on AI for Prosperity published in June 2025, has stayed with me:
“We will drive economic growth, address talent shortages, and ensure equal opportunity, by encouraging girls…to pursue science, technology, engineering, and mathematics (STEM) education and increasing women’s representation in the AI talent pool at all levels.”
photo: THE CANADIAN PRESS/Adrian Wyld
(a similar statement was made at the 2018 G7, which I attended.)
Years of experience in technology suggest that statements like these, while well-intentioned, rarely translate into meaningful change. The lack of women in STEM is not due to a lack of ability, interest, or encouragement. It persists because existing structures protect those who hold power. Too often, that power is unconsciously guarded by those in authority, including professors, employers, or colleagues, who shape who gets access, recognition, and advancement.
Without bold, risk-taking plans that shift ownership and redistribute opportunity, statements like the G7’s remain symbolic. History does not simply repeat; it compounds. Each time we fail to act, inequity deepens.
Power has always followed knowledge. During the first two industrial revolutions, reading, writing, and numeracy controlled information and progress. In the digital revolution, coding created the digital world (along with a lot of money for those writing it). Across these revolutions, power was predominantly held by white men.
Now, in the age of artificial intelligence, power belongs to those who own and shape data. Algorithms move words, images, numbers, and videos across the physical and digital worlds. Control over data means control over knowledge and, again, over power itself.
By the mid-1990s, the gender gap in STEM was already a topic of discussion. By 2000, as the term STEM gained traction, programmers, overwhelmingly white and male, were designing a digital world that reflected existing social and economic hierarchies.
Fast forward to 2025, and the gap persists. A UNESCO report shows women make up only 26% of the global data and AI workforce, 15% in engineering, and 12% in cloud computing (UNESCO). These figures have remained largely unchanged for decades. Despite billions invested in “girls in STEM” initiatives, the system has not shifted. Participation at the entry level has improved, but the deeper structures of power remain intact.
Real change, the kind that reshapes societies, requires more than participation programs. It demands a redesign of the systems that concentrate power and control knowledge. Addressing the STEM gender gap requires democratising power, not just increasing participation.
In education, every young person must be prepared to see and challenge bias, to work together as equals, and to understand how data and AI influence the world. Classrooms should not only teach algorithms, but also encourage students to ask who writes them and for whom.
In leadership, systems that reward only technical skill or narrow definitions of merit must evolve to value inclusion and accountability. True leadership requires the courage to redefine excellence itself.
In policy, accountability must be measurable. Institutions should be evaluated on outcomes, with funding and accreditation tied directly to evidence of equitable structures and results.
For years, the focus has been on teaching girls to code, an important but incomplete effort. We have not yet taught boys and men, many of whom are bosses, professors, and senior leaders, how to create environments that truly inspire, engage, include girls and women, and how to share the power. (On a side note, my next blog will highlight three women mentored over the past decade in STEM, illustrating the challenges they faced from high school to the workforce.)
As AI increasingly shapes our lives, these inequities risk becoming entrenched. AI is not neutral; it learns from our data, histories, and biases and amplifies them at scale. Without intervention, tools designed to advance society may instead reinforce existing divides.
The Change Must Be Systemic
Closing the STEM and AI gender gap is not only about representation; it is about preventing bias from being built into these systems. Gender equity must be integrated into every stage of development, from education to leadership to algorithm design.
Education is central. When students understand how technology reflects human choices, they see that AI mirrors the values of its creators. Embedding AI and data literacy, with ethics, into curricula fosters critical awareness and responsibility. Teachers need training and resources to teach inclusively and help students recognise and question bias in digital tools.
In higher education, universities should publish gender data on STEM enrolment, retention, and advancement, linking funding to progress. Faculty and research leaders should receive AI ethics training to ensure future developers design fair systems.
Workplaces must meet similar standards. Public funding for AI projects should require diverse leadership, inclusive hiring, and ethics training. These measures help teams understand how algorithmic decisions can reproduce inequality.
What a Bold Plan Could Include
Gender-Impact Audits: Publicly funded AI or data initiatives must assess effects across design, development, and deployment. Projects that fail to address widening gender gaps will face corrective action, including possible funding penalties.
AI Literacy & Ethics Training: Executives, policymakers, and technical leads must complete mandatory AI literacy and ethics training focused on gender-aware governance. Certification should be required for managing publicly funded projects.
Link Policy to Accountability and Funding: Government contracts, research grants, and subsidies should be tied to measurable progress on gender balance and inclusion. Non-compliance will result in funding suspension or public reporting.
Transparent Reporting and Oversight: Annual reports should publish disaggregated data on gender representation, retention, and pay equity. An independent oversight body should monitor compliance, evaluate audits, and enforce accountability.
Curriculum and Workforce Reform: AI, data, and ethics education should be embedded across K–12 and higher education, with teacher training grounded in equity-centred design. Targeted scholarships, mentorship, and re-entry programs will support gender diversity.
Equity-by-Design in AI Systems: Gender equity must be embedded in every stage of AI system design, development, and deployment, including diverse teams, bias detection, and ethical impact assessments.
If we want equity in STEM, we need bold, evidence-based policies that reshape education and governance so that power is shared, knowledge is democratised, and the systems of the future reflect the diversity of the societies they serve.
As Canada and countries worldwide roll out AI strategies, we have a rare opportunity to take systemic action. AI is advancing at full speed; we cannot afford the next G7 statement to repeat again “…encourage girls and women into STEM fields.”
Sources
UNESCO Flags Gender Gap in STEM – Education Today, May 2025
Women Make Up 29% of the AI workforce – Here’s How to Fix it – Forbes, Nov 2024
Can AI fix the gender gap in STEM? World Economic Forum, 2025
UNESCO 2024 Gender Report – Global Education Monitoring Report
Why women’s reluctance to embrace AI could be a career killer – Globe & Mail, July 2025
Girls Unstoppable – A celebration of a brighter future for girls – Lego, 2024
Gender differences in STEM enrolment and graduation: What are the roles of academic performance and preparation? – Statistics Canada, 2021
Gender differences in Science, technology, engineering, mathematics, and computer science programs at university – Statistics Canada
Leave a comment