AI Tells Equally Qualified Women to Ask for Lower Pay, White Paper Warns

Artificial intelligence is advising women to ask for much lower salaries than men with the same educational qualifications, skills and experience for the same job, a study conducted in 2025 has found. According to a report published by the technology website The Next Web, several AI models used in the test consistently told women to claim lower pay in salary negotiations, with the gap reaching about 10 percent in many cases.
The UK women's rights organisation the Fawcett Society has recently published a white paper titled 'Can We Afford to Automate Gender Inequality?'. It shows how AI-based systems can deepen existing gender bias in hiring, pay and promotion.
The white paper's author, Mariyam Yakub, said: 'Artificial intelligence does not create discrimination from nothing. It learns from the bias that already exists in our institutions and repeats it on a large scale. If we do not act now, the institutional systems that already discriminate against women will simply be converted into automated systems.'
According to the report, by 2030 up to 40 percent of women in some industries could be at risk of losing their jobs because of automation technology and AI. Analysis by the International Labour Organization (ILO) has also found that women-dominated professions face far more change and risk from artificial intelligence than male-dominated professions.
The report cites Amazon as a widely discussed example of bias in hiring. The company had experimentally built an AI-based recruitment system trained on past hiring data. The technology learned to give more weight to male candidates, and women's applications were being rated comparatively lower. Amazon eventually scrapped the system.
The report says AI bias differs from human bias in three ways — scale, speed and institutionalisation. One employer's bias may harm a few women, but a biased AI system can make discriminatory decisions about thousands of candidates in an instant. Because technology-based decisions are widely assumed to be neutral and objective, identifying the discrimination and challenging the decision also becomes difficult.
For Bangladesh, the report cites the World Economic Forum's Global Gender Gap Report 2026. On economic participation and opportunity, Bangladesh ranks 142nd among 145 countries. While 96.1 percent of the education gender gap has been closed, the figure for the economic field is only 41.9 percent. Women make up just 6.2 percent of legislators, senior officials and managers.
The report calls for mandatory gender-impact assessment and regular audits before AI used in hiring, pay and promotion is deployed. It also recommends that training data be representative, that workers be told whether AI was used in a decision about them, that there be a right to object to discriminatory decisions and a human review mechanism, that the government create effective policies and an accountability framework for high-risk AI use, and that women's opportunities to gain AI-related skills be expanded.



