If we get it right, AI could free women from repetitive tasks and open doors to better careers. If we get it wrong, we will look back in ten years and wonder how a technology that promised so much pushed so many women backwards
Walk into almost any bank branch in Lagos or Abuja and look at who is doing the work. Look at the tellers, the customer service desk, the account officers and the staff processing documents in the back office. Call a telecoms helpline and listen to the voice that answers. Visit the administrative floor of a ministry or a large company. In many of these places, you will find women.
These are also the jobs that artificial intelligence is changing first.
This year, the International Labour Organisation studied how generative AI affects different kinds of work. It found that jobs held mostly by women are almost twice as exposed to the technology as jobs held mostly by men, 29 per cent compared with 16 per cent. Clerical and administrative roles face the highest exposure of all.
The reason is simple. Generative AI is good at the tasks that fill much of the work in these jobs: answering routine questions, drafting letters, entering data, summarising documents and checking forms.
The World Economic Forum’s Global Gender Gap Report 2026, released last week, reached a similar conclusion. It warned that AI could widen gender gaps at work because the roles it is disrupting are the ones where women are concentrated.
Exposure does not always mean job loss. Sometimes AI takes away the boring parts of a job and leaves the more valuable parts behind. But whether that happens depends on decisions employers make, and those decisions rarely favour the people at the bottom of the organisation chart.
I spent more than twenty years working on large technology projects in the energy sector. One pattern repeated itself on almost every programme. When a new system arrived, the people whose daily work changed the most were usually the last to be consulted and the least prepared. They worked in operations, finance support and administration. Many of them were women.
AI threatens to repeat that pattern on a much larger scale.
Think about what happens when a Nigerian bank introduces a chatbot capable of answering most customer questions. The bank saves money. The technology team gets praised. The customer service staff are told to upskill. But upskill into what, paid for by whom, and in what time? A woman working long shifts at a contact centre, then caring for children and relatives at home, does not have free evenings for online courses. If retraining is left to personal effort, many women will fall behind.
The risk lands on the people with the least power to shape the change.
Now look at the other side. AI is also creating work, and much of it is well paid. Employers want people who can build AI tools, manage data, protect systems from cyber-attacks and redesign how businesses run around the new technology.
Women are not getting enough of these roles. The same World Economic Forum report shows that women make up only about one in five AI engineers. In AI companies, women hold about a quarter of top management positions, fewer than in comparable firms outside the sector. Even among founders, 15 per cent of men list AI skills on their profiles, compared with 9 per cent of women.
Put these findings side by side and a worrying picture appears. AI is changing the jobs women hold today, while the new, better-paid jobs it creates are going mostly to men. If nothing changes, AI will not just maintain the gender gap. It will make it wider.
This is not a future problem.
It is tempting to treat all of this as something that will happen later. It is happening now. Nigerian banks, telecoms companies and outsourcing firms are already testing and using AI tools. Decisions about which roles to cut, which to redesign and who gets trained for the new ones are being made in boardrooms this year.
That is why the conversation about women and AI must move beyond celebration. Every time we applaud a woman who has built an AI startup, and we should, we must also ask about the thousands of women whose jobs are being quietly redesigned without them in the room.
First, employers should map out which roles AI will change before introducing it, and share that plan with staff. Workers deserve to know what is coming.
Second, retraining should happen during working hours and be paid for by the organisation. Moving an experienced customer service officer into a role supervising AI systems, handling complex complaints, or analysing customer data costs less than hiring from outside and keeps valuable knowledge within the company.
Third, the government should track who is affected. Nigeria’s National AI Strategy aims for half of those trained in AI skills to be women. That is welcome, but we also need to know whether those women end up in better jobs, not just whether they attended a course.
Fourth, companies should report on gender in their AI teams, the same way many now report on gender on their boards. What gets measured tends to get fixed.
AI will not decide on its own who wins and who loses. People will. Bank executives, company boards, regulators and training providers are making those choices now.
So the question is not whether AI will change women’s work. It already is. The question is whether the women who carry the risk of this change will also get a fair share of its rewards.
If we get it right, AI could free women from repetitive tasks and open doors to better careers. If we get it wrong, we will look back in ten years and wonder how a technology that promised so much pushed so many women backwards.
*Motunrayo Opayinka (Mrs) is a technology and business transformation executive with over 20 years of experience using technology to solve complex problems and create opportunities. As Founder of Womenovate, she champions women, young people and underserved communities in the digital economy.
