Today’s headlines in artificial intelligence (AI) are dominated by a few famous founders, CEOs and wunderkinds – mostly men. Yet, built upon disciplined engineering, thoughtful product decisions and patient infrastructure planning, women’s contributions to the AI industry have been just as profound.
Architects of AI
The reality is that women have been among the most influential architects, shaping AI domains that are critical to its growth and advancement. Some of the most enduring influences in AI are almost invisible, a bit like good architecture. Take Fei-Fei Li, widely considered the “godmother” of AI. AI was a niche research domain with many small datasets and inconsistent evaluation before computer vision became mainstream. But her 2009 ImageNet paper, which introduced a large hierarchical database of labelled images, made it feasible to train and compare AI models on a shared reference point, providing a solid foundation for cumulative, rather than fragmented, progress.
While Sam Altman received most of the press attention when generative AI (GenAI) started to scale commercially, it was Mira Murati who led major initiatives including ChatGPT and DALL-E as the Chief Technology Officer. She built the company’s teams and platforms, established its engineering discipline and shaped its product trade-offs and design choices. She subsequently launched Thinking Machines Lab, a company dedicated to making AI more understandable and relatable to ordinary people and organisations, and enabling them to steer toward their own goals.
Advocates for Zero Bias
Another way women have reshaped AI is by pushing for the measurement of objective reality, instead of relying solely on performance. In “Gender Shades” published by Joy Buolamwini and Timnit Gebru, they highlighted how large disparities in commercial classification across intersections of gender and skin type can create the illusion of accuracy. They then went further in “Datasheets for Datasets” by proposing standardised documentation describing datasets’ motivation, composition, collection process, recommended uses and limitations. Their work underpins responsible AI – something that seems obvious today, but wasn’t before, when datasets carried hidden assumptions.
Guardians of Trust and Privacy
Concurrently, Cynthia Dwork’s work on differential privacy formalised a rigorous way to extract insights from data while limiting what can be inferred about any individual. This concept is increasingly important as AI moves deeper into domains of health, finance and personal communication. Similarly, Barbara Grosz’s advocacy for designing collaboration into intelligent systems from the start provides an early blueprint for the assistants and agentic tools that are now becoming mainstream.
Masters of Precision
Finally, women like Regina Barzilay have shaped AI by emphasising that precision is not a nice-to-have in high-stakes fields. Through her evaluation and validation of Mirai, a mammography-based breast cancer risk model, across diverse hospital systems, she ensures it performs reliably across populations, sites and workflows. The future belongs to systems that are not only impressive but robust.
In 1972, Karen Spärck Jones introduced the term-specificity idea, now known as Inverse-Document Frequency (IDF), which is how search engines find and prioritise information. Even in an era of large language models, her work remains very much relevant.
A unifying thread runs through these stories – the contributions of women in AI are no less impactful than men’s, even if they are not always immediately visible. Their work gives us hope that AI will ultimately boost human development, rather than lead us toward a dystopian future.
First appeared in the IT Society Magazine from the Singapore Computer Society
