Gendered Leadership Agency and Generative AI: A Conceptual Note from Academic perspective

In the modern university, leadership is not merely about issuing orders or managing budgets; it is a deeply human practice rooted in relationships, ethical judgment, and the daily negotiation of complex social expectations. For decades, researchers have understood that these roles are not gender-neutral. Women in academia often find themselves shouldering a disproportionate share of "academic housework"—tasks like mentoring, emotional support, and administrative coordination—while traditional ideals of leadership still favor masculine traits like individual authority and decisive action. Now, a new force has entered this landscape: generative artificial intelligence. Unlike earlier computer tools that simply performed calculations, this technology can write text, generate images, and synthesize information, acting as a partner in the very act of thinking. As these tools become woven into the daily work of university leaders, a critical question arises: how does this technology change the way women lead, and does it help or hinder their ability to exercise judgment and authority?

A team of researchers from Tampere University in Finland set out to explore this question by listening directly to the people living through it. They conducted in-depth interviews with eight female academic leaders, half from the Global North and half from the Global South, all of whom were actively using generative AI in their work. Rather than trying to prove a universal rule or predict the future, the authors offered a conceptual framework to understand what is happening right now. Their work suggests that the arrival of these intelligent tools is quietly reshaping the very architecture of leadership, shifting where autonomy and responsibility reside in ways that are often invisible.

The most striking discovery from these conversations is that agency—the capacity to act and make choices—is being redistributed. In the past, a leader's agency was clearly located in their own mind and decisions. Today, it is a negotiation between the human and the machine. The women in the study described a careful, ongoing boundary work. They might use the AI to draft a speech or polish a document because it is faster and more fluent, yet they insist on writing a version themselves first to preserve their "thinking sovereignty." They use the tool selectively, refusing to let it generate full texts, ensuring that the final judgment remains human. This is not a simple rejection of technology, but a constant, active effort to decide what parts of the work can be offloaded and what parts must remain strictly human.

However, this shift brings a new and troubling complication regarding bias. In traditional university settings, gender bias is often visible and nameable. A woman might be passed over for a promotion because a committee prefers a man, or she might be bypassed in meetings because colleagues unconsciously look to her male subordinates for answers. These are human errors, rooted in specific people or institutions, and they can be challenged. The researchers found that generative AI changes the nature of this bias. When the AI produces an image of a professor that is overwhelmingly a white man, or when its language defaults to male pronouns, the bias is no longer attached to a specific person or a decision-making committee. Instead, the bias is embedded in the data itself, a diffuse and unattributable force. There is no one to hold accountable, no specific role to contest, making the prejudice harder to see and even harder to fight.

This leads to a second major finding: the emergence of a new kind of leadership skill that the authors call "critical AI literacy." This is not merely the technical ability to use the software or the caution to avoid it. It is a form of ethical and relational judgment. The leaders in the study realized that they must do more than ask if the AI's output is correct; they must ask if it is right, and for whom. They must possess the capacity to recognize when a tool is reproducing harmful stereotypes or reinforcing existing power structures, even when the text looks perfect. This literacy requires a deep engagement with the technology to understand its limitations, rather than a fearful disengagement. One leader noted that those who refuse to use the tool are likely to develop stereotypes about it, while another emphasized that only first-hand experience allows one to move beyond fear-based narratives.

The researchers highlight a profound paradox at the heart of this transition. As leaders rely more on AI for cognitively demanding tasks like writing and analysis, they risk eroding the very judgment and interpretive capacities that leadership requires. The more the machine does the thinking, the more the human leader must work to maintain their own ability to think critically. This is particularly urgent for women leaders, who already navigate a landscape where their authority is often questioned. The study suggests that the future of academic leadership will not be a choice between human or machine, but a co-mediated practice where humans and algorithms work together. Success in this new environment will depend on the ability to navigate the space between them, maintaining human sovereignty while recognizing that the tools we use carry the hidden assumptions of the data they were built upon.

Ultimately, this work is a call for further investigation. The authors caution against assuming that the relationship between gender and technology is simple or that the experiences of these eight women represent every leader. They argue that we need to understand how gender and agency are being co-constructed with AI in everyday life, moving beyond old ideas of leadership that treat technology as a separate tool. The goal is not to ban these technologies, but to ensure that as they become central to how we work, we do not lose the human capacity for ethical judgment, care, and the ability to see the world clearly.