It is 2025. International Women's Day celebrates women and highlights much of the inequality women face. Yet, to make real progress in technology and an increasingly digitalised world, we need to look beyond general gender divides. Showing diversity data by gender masks age composition.
It hides how women disappear from the workplace due to ageism, sometimes voluntarily, but often in a more subtle way due to negative behaviour. The problem is that this negativity is increasingly determined and reinforced by the design and use of AI in the workplace.
Are older women simply beneficiaries, not contributors to AI?
The age at which ageism starts to impact women varies by country, economic factors, and cultural factors. These topics are already well covered elsewhere. My question is not about how society looks after this group of women.
Instead, it is the more fundamental issue of how older women's value in the workplace is changing with AI and corporate algorithms. Why do we need to prevent actions that negatively influence decision-making towards this group?
There is little research on this topic, with articles focusing on how AI helps women. In Google search, AI explains how older women can benefit in the workplace. It says, "as a tool to enhance their skills, overcome potential age-related biases, and remain competitive in the job market by automating repetitive tasks… allowing them to focus on higher-level strategic thinking, provided they receive proper training and support to utilise the technology effectively."
Some research fails to share data by gender. For example, McKinsey's study "AI in the workplace: A report for 2025 "gushes about how young people benefit from AI. It says, "C-suite leaders can help millennials lead the way," while millennials (aged 35 to 44) are described as "having the most experience and enthusiasm about AI, making them natural champions of transformational change." However, these claims do not provide details on gender or include older age groups.
Older women are receiving less AI training.
Why does AI need diversity, and does it benefit from the experience of older women? Or are the skillsets of older women now redundant? AI in Google search has no answer to this question, but research by other organisations has provided some ideas.
The adoption of AI in the workplace has forced everybody to learn new skills, whether young graduates or baby boomers. Yet, access to learning these new skills is not equal. A recent study by Randstad highlighted the issue of providing AI training.
Their study showed that while 75% of companies are adopting AI, only 35% of talent had received AI training in the last year, with a male/female split of 71%/29% for AI-skilled workers. This is not surprising, but lack of opportunity is a more significant concern.
Only 22% of Baby Boomers were offered AI skills training compared with 28% of Gen X and 45% of Gen Z workers. So, older workers are at a disadvantage in maintaining relevant skills. In Randstad's report, women working for around 30 years account for only 21% of talent with AI skills; in Generative AI, this is 24%, and in Deep Learning, this is 15%.
Furthermore, this report highlighted how this same group is underrepresented in specialist AI skills such as software development (18%), AI data processing (19%), AI cloud applications (20%), applied machine learning (23%), deep learning (24%) and generative AI (31%).
Why are Older Women Critical to AI's development?
Research shows that most people in the workforce have only one year of experience with AI (five years of experience is already considered high). So, with comparable amounts of AI experience to other age groups, what additional value does older women have in a world of fast-changing technology?
Women with over thirty years of experience bring deep knowledge of other sectors, different behavioural patterns, and cognitive responses to situations compared with their younger peers. These responses and their differences are precisely what are needed to make AI more accurate, useful, and valuable to society.
This is no secret. IBM has also highlighted the challenge of AI development. It notes, "a tiny homogeneous group of people determine what data to use to train generative AI models, which is drawn from sources that greatly overrepresent English." Empirical evidence demonstrates that all predictive models, including AI, are more accurate when they incorporate diverse human intelligence and experience.
The Diversity Prediction Theorem explains this. The theorem demonstrates how the more significant the diversity, the better the prediction. This also lowers the likelihood of AI hallucinations caused by perceived patterns.
How can we recruit older women?
Recognising that older women can add value, we need to ensure they are included rather than excluded in recruitment processes, whether internal or external. In order to create an inclusive workplace, the design of technology, including algorithms, must be reviewed. This will ensure that this group is not frozen out (even inadvertently) by the limited experience of young designers.
Conclusion
The employment of older women can create massive value for a workplace. Diversity of views is critical for AI's sustainable, healthy development, and all workplaces can benefit from cross-industry experience.
Finally, it is worth remembering that older women also make good employees because they typically bring stability and commitment. Many of the concerns about employing younger women are less relevant.
Most older women won't have personally benefited from many of the employment policies that have come into play in recent years. Therefore, their expectations of a workplace are generally lower, making them less demanding.
Not too old, but highly experienced! Older women can add value to AI!
