AI Bias May Worsen Gender Inequality in Jobs: ILO Report
Since AI models are trained on historical data that often reflects male-dominated patterns, they risk reinforcing discrimination in hiring, pay, and access to financial services
Chennai: Artificial intelligence can disproportionately affect women’s jobs, and AI systems trained on biased and incomplete data can deepen gender biases in the workplace, a report by the International Labour Organization (ILO) has found. In India, awareness of this threat remains nascent and fragmented.
A recently released ILO report reveals a concerning trend: jobs dominated by women are nearly twice as likely to be exposed to generative AI compared to male-dominated roles. Around 29 per cent of female-dominated occupations are exposed to GenAI, compared to just 16 per cent of male-dominated occupations.
The report also finds that systems trained on biased and incomplete data can disadvantage women in recruitment, pay decisions, credit scoring, and access to other services.
Since AI models are trained on historical data that often reflects male-dominated patterns, they risk reinforcing discrimination in hiring, pay, and access to financial services. In India, this issue is compounded by women’s lower digital and financial footprints and the invisibility of informal work, leading to incomplete datasets, said Rituparna Chakraborty, Partner and Regional Lead for India at True Search.
Awareness about gender bias in AI remains limited and fragmented.
“AI will not create gender inequality, but it can either correct it or compound it. The choice India makes now will determine whether AI becomes an equaliser or an amplifier of bias,” she said.
She stressed that the first step is acknowledgement and awareness, followed by willingness and clear intent to address the issue.
From a government perspective, mandated algorithmic audits for bias, similar to ESG compliance, are needed. “Women can be incentivised to participate in AI skilling programmes at scale. Gender-tagged datasets can be developed through public digital infrastructure,” she said.
From an industry perspective, she suggested embedding gender considerations into AI systems by design and regularly auditing hiring, pay, and promotion algorithms.
She also emphasised the need for returnship programmes for women in technology and greater inclusion of women workers in AI training datasets and testing processes. A multi-stakeholder dialogue involving government, industry, and academia is essential to address emerging risks.
Globally, roles such as clerical work, administrative support, customer service, data entry, and BPO operations are most vulnerable. In India, this risk is even more pronounced due to the large number of women employed in these repetitive, task-heavy sectors, especially in the outsourcing industry.
Even now, a large proportion of working women are concentrated in agriculture, often in informal or unpaid roles. Others are employed in sectors such as healthcare, education, retail, and domestic work. In white-collar jobs, women are mainly seen in entry- and mid-level roles in IT, HR, and finance, with limited representation in leadership positions, she said.
