Why AI Still Isn’t Working for Women

This International Women’s Day, conversations across the technology industry are focusing not only on progress for women in tech but also on the role artificial intelligence plays in shaping their safety, representation and opportunities.

In recent months, several headlines have raised serious concerns about how AI is being developed and deployed. These include the use of AI to generate millions of non consensual intimate images of women late last year, growing fears around women’s safety linked to facial recognition smart glasses being developed by Meta, and criticism aimed at a recent “AI Skills 4 Women” training course by Microsoft that was accused of reinforcing stereotypes and lowering expectations for women entering the field.

While these incidents might appear unrelated, industry leaders say they reveal a much deeper and more systemic problem within AI development.

AI Bias and Gender Inequality in Technology

Maisie Holder, COO at World Summit AI, believes the recent controversies surrounding AI and women are not accidental.

“The recent patterns and events we’ve seen taking place within tech, and specifically within AI, haven’t arrived by chance,” Holder says.

“AI models and technologies behave as they’re trained to behave. They reflect the biases and blind spots present in the environments in which they are developed.”

At the centre of many criticisms of AI systems is a recurring issue.

“Present throughout all the current criticisms of AI in relation to women is a common theme: exclusion. Despite progress compared to decades past, women remain underrepresented in the governance, development, and investment decision making processes related to technology and AI.”

“When diverse voices are missing from the room, products lacking fair representation result, which can reinforce harm, even inadvertently.”

Holder stresses that addressing these issues does not mean slowing innovation. Instead, it means recognising the influence AI now has on daily life.

“The goal here isn’t to slow down development or innovation. It’s about encouraging recognition that AI is now deeply embedded in the systems that shape our environments and, consequently, our safety.”

“It’s a huge focus for us at World Summit AI, and why we ensure there is equal representation throughout our agenda. If we fail at addressing structural inequality at the core, we risk AI scaling and growing with those inequalities irreversibly baked in.”

Why Diversity in AI Development Matters

Sarah Porter, founder of InspiredMinds and World Summit AI, says that acknowledging the problem is only the first step.

“These stories and headlines raise fair critiques of some of the ways AI is being utilised and rolled out, and that’s important,” Porter says.

“However, it can’t stop there. The important next step is a move towards accountability and action. Decision makers and developers need to embrace the challenge of improving inclusion. It isn’t a voluntary upgrade; it’s an opportunity to build technologies better to cater to and protect half the human population.”

To explain the risks of ignoring this issue, Porter points to the history of medical research.

“In the medical field, it wasn’t until 1993 that women were first included in early phase clinical trials. As a result, accurate dosing and understanding treatment side effects have historically been a huge issue for women.”

“AI development is on a similar precipice. If we go too long without training AI with diversity and safety for women at the forefront of its development, the damage will extend into the years ahead and cause real harm.”

Building Safer and More Inclusive AI

Porter believes the solution is clear, though it requires commitment across the industry.

“Diverse technical teams are a minimum requirement. Comprehensive bias and impact assessments at regular stages of development are better still.”

“In the event that products are released or upgrades are deployed that place women at risk, it’s vital that the industry engage transparently to address the issue.”

“It isn’t the time for defensive behaviour or evading accountability.”

close

As artificial intelligence continues to shape everything from digital platforms to everyday devices, the decisions made today will define how fair and safe these systems are for years to come. On International Women’s Day, the message from many across the AI community is simple. Technology that is built for the future must be built with everyone in mind.