Breaking the algorithmic glass ceiling: How AI can shape a more inclusive future for women leaders

Artificial intelligence is rapidly becoming embedded in the systems that determine who gets hired, who gets promoted, and who is identified as future leadership material. What was once shaped primarily by human judgement is increasingly influenced and, in some cases, determined by algorithmic recommendation.

As someone who has spent my career in technology, I am optimistic about AI's potential. Used responsibly, it can improve consistency, reduce individual bias and unlock productivity. But I am equally conscious of a growing challenge: if we are not deliberate, AI could reinforce inequities that many organisations have worked hard to address.

In doing so, it risks creating a new and less visible barrier for women aspiring to leadership.

Women remain underrepresented in senior technology roles globally, holding less than a third of tech positions and an even smaller proportion of executive roles. At the same time, AI systems are being trained on historical workforce data that reflects decades of uneven access to opportunity.

When these realities intersect, the consequences warrant attention.

When historical data shapes future leadership

AI systems learn from patterns in past decisions. If historical hiring and promotion trends favoured linear career paths, uninterrupted tenure or certain leadership styles, algorithms trained on that data may interpret those patterns as indicators of success.

This is not intentional discrimination. It is statistical inheritance.

For example, recruitment tools that prioritise specific career trajectories may undervalue non-linear paths or career breaks and experiences that are more common among women. Performance management systems that emphasise constant visibility, responsiveness or digital presence may disadvantage those working flexibly. Leadership prediction models can inadvertently reward familiarity, selecting candidates who resemble those who have historically held power.

The impact of any one decision may appear minor. But AI operates at scale. A biased hiring manager might influence dozens of outcomes. A biased algorithm can influence thousands.

Small skews in shortlisting, performance scoring or high-potential identification compound over time. Those early filters shape access to stretch assignments, transformation programs and succession planning pools. And leadership pipelines are built years before executive appointments are made.

These patterns, if unaddressed, can unintentionally limit opportunity over time.

The governance opportunity we must embrace 

The greater concern is not that bias exists; it is that AI is often perceived as inherently objective.

When decisions are labelled "data-driven," they can become harder to question. Yet data reflects the priorities, structures and inequalities of the environments in which it was generated. Without oversight, organisations may normalise inequitable outcomes under the banner of efficiency.

This is where governance becomes critical.

Women remain underrepresented in advanced AI development roles and in executive forums overseeing digital transformation. The questions that shape AI systems - What datasets are we using? How do we define success? How are models tested for disparate impact? - are technical, but they are also strategic.

Diverse oversight is not symbolic. It is a risk management imperative.

AI is also reshaping work itself. As automation reduces routine tasks, new high-value roles are emerging in areas such as data governance, AI ethics, digital strategy and transformation leadership. These roles are increasingly pathways into senior executive positions.

If women are not proportionately included in reskilling initiatives or assigned to transformation programs, the leadership gap of tomorrow will widen - not because of capability, but because of access.

The World Economic Forum estimates that automation will significantly transform a large proportion of current job roles within the next five years. The redistribution of opportunity is already underway. Ensuring equitable participation in that shift is not a diversity initiative alone; it is a workforce strategy.

So, what should organisations do?

First, any AI system influencing hiring, promotion or workforce restructuring should undergo regular independent audits for bias and disparate impact. Outcomes should be monitored by gender and other diversity indicators not only for accuracy, but for fairness.

Second, procurement standards must require transparency from technology vendors regarding training data, model validation and explainability. Responsible AI cannot operate as a black box.

Third, leadership potential metrics should be broadened. If models are trained on narrow proxies of past success, organisations must redefine what they value - recognising adaptability, cross-functional leadership and learning agility alongside traditional indicators.

Finally, AI governance should sit at the executive and board level. Oversight committees must include diverse voices and clear accountability structures. Responsible AI is not solely an IT issue; it is an organisational one 

As AI becomes embedded in hiring, promotion, and leadership development systems, we have an opportunity to ensure it reflects the inclusive workplaces we aspire to build. Technology does not determine outcomes on its own. The choices we make about how it is designed, governed, and applied will shape its impact 

If inclusion is intentionally embedded into AI systems, they can help expand access to opportunity, surface diverse talent and create more consistent decision-making. If we approach AI with discipline, transparency and accountability, it can strengthen rather than narrow leadership pipelines.

Innovation is not only about adopting new tools. It is about applying them responsibly and thoughtfully.

The leadership teams of the future are being shaped today through the data we prioritise, the metrics we value and the oversight structures we establish. Ensuring those systems promote fairness and broaden access is not a diversity initiative alone. It is a leadership responsibility. 

On International Women's Day, the question is not whether AI will influence the future of work. It is how we will guide it  so that progress for women continues not only in policy and intention, but in the systems that underpin opportunity itself.