Introduction

Advancements in artificial intelligence (AI) have the potential to reshape multiple sectors, but their impact on existing social inequalities is increasingly under scrutiny. Notably, recent findings suggest that AI technologies can perpetuate historical gender biases, particularly regarding leadership representation.

Gender Representation in AI-generated Content

In a recent experience, a request for AI-generated images to illustrate leadership resulted in a skewed representation. Initial attempts yielded images predominantly featuring male leaders, with only specific prompts leading to depictions of women in leadership roles. This illustrates a concerning trend in AI: the predominant portrayal of male leaders as amicable and engaging, while images of women often depict them alone, contributing to a narrative of leadership as a predominantly male domain.

The Evidence of Bias

Research conducted by the Berkeley Haas Center for Equity, Gender, and Leadership highlights that approximately 44% of AI systems across various industries exhibit gender biases. A significant number of these systems also reflect both gender and racial biases. Furthermore, findings from UNESCO reveal that large language models, especially those available for free like Llama 2 and GPT-2, tend to represent women in stereotypical roles. Notably, women are described in domestic contexts far more often than men, reinforcing traditional gender stereotypes.

Environmental Impact and Inequities

The environmental implications of AI technologies are another layer in this issue. Training AI models requires substantial energy and resources, leading to significant carbon emissions—up to five times the lifetime emissions of an average car for a single model. This environmental strain disproportionately impacts regions confronting climate stress, often worsening conditions for women, who frequently bear the responsibility of managing household resources.

A Call for Inclusive AI Development

For AI to be a tool that genuinely benefits all, it must be developed inclusively and sustainably. Addressing foundational inequities—such as the representation of women in media, politics, and the technology sector—is essential. Currently, women represent only 30% of the AI workforce, with a significant gap in leadership positions, exemplified by a 64% gender gap in India.

“We cannot afford AI to prosper at the cost of women’s lives.”

Conclusion

While AI holds promise for innovation and progress, its development must prioritize inclusivity and environmental sustainability. It’s imperative to encourage women to participate actively and equitably in the technology sector, particularly in leadership roles, to counteract long-standing biases and create a more balanced future.