Tech Explained: From code to care: Making gender central to India’s AI future in Simple Terms

As India prepares to host the India AI Impact Summit (from February 16 to February 20), the country stands at an inflection point. With ambitious projections of a $500-billion artificial intelligence (AI)-driven economy and aspirations of becoming a global AI hub, India’s technology narrative is firmly centred on growth, productivity, and innovation. However, one critical question that remains underexplored is whether India’s women will be able to equally participate in shaping, benefiting from, and being protected within this AI ecosystem.

If India is to lead responsibly, gender must not be relegated to symbolic representation at panel discussions but be embedded at the core of AI policy, design, deployment, and governance.

AI is not gender neutral
India’s National Strategy for Artificial Intelligence envisions AI as a driver of inclusive growth. But technological systems are not inherently inclusive; in fact, they often amplify the inequalities embedded in the societies that produce them. India’s female labour force participation has historically lagged that of men; consequently, women remain under-represented in high-value technology roles. Globally, women constitute less than a third of AI professionals. In India, they make up roughly one-fifth of the digital workforce and an even smaller share in AI research, advanced technical roles, and leadership positions.

When women are absent from design teams and decision-making tables, the technologies built reflect narrower perspectives. For instance, international evaluations of facial recognition systems have shown higher error rates for women, particularly darker-skinned women. AI-based hiring tools have reproduced historical gender bias, rather than correcting for it. Beyond just technology access, value-added services like credit scoring systems often penalise career breaks linked to caregiving. Representation in AI is, therefore, not simply about jobs but is about fairness, accuracy, and democratic accountability.

Building Indian AI without importing global bias
Much of today’s mainstream AI has been trained predominantly on Western, English-language, male-dominated datasets. These knowledge systems often under-represent women’s experiences, unpaid care work, informal labour, caste realities and rural economies—all central to India’s social fabric. When such models are deployed without contextual correction, they reproduce skewed assumptions, such as default male professionals, the absence of caregiving, and a limited representation of marginalised communities.
If India simply localises international AI architectures without changing their biased foundations, it risks replicating the very exclusions it seeks to overcome. An “Indian AI” must not merely be linguistically Indian but be socially representative. This requires investing in diverse and gender-balanced datasets; ensuring women are part of AI design teams; and recognising unpaid and informal work on economic datasets.Acknowledging time poverty, digital gaps, and trust deficits
A deeper constraint lies in women’s “time poverty”. India’s Time Use Survey 2025 shows that women spend nearly three times as much time as men on unpaid domestic and caregiving work. This translates to working women spending around 10 hours less than men every week on self-development activities, including learning and skill enhancement. This time gap happens when women are in their prime (25-45 years), handicapping their ability to transition from low-skilled to higher-value work in an AI-disrupted economy.

Unequal digital access compounds this problem. Rural women are significantly less likely to own smartphones or have reliable internet connectivity compared to men. Moreover, women express higher levels of distrust toward AI systems, citing concerns around bias, privacy, and online safety. The rapid rise of AI-generated deepfakes and online harassment disproportionately targeting women leaders reinforces this anxiety. Therefore, an AI ecosystem that does not proactively address safety, trust and access will struggle to achieve meaningful female participation.

The economic case for inclusion
About 50% of men compared to 37% of women have reportedly adopted generative AI, implying that AI ecosystems are male-dominated, even at the user level. If women are excluded from advanced skilling or perceived as less technically adept, a “competence penalty” may emerge that will widen wage gaps and reinforce occupational segregation.

Many routine administrative and clerical roles, sectors where women are over-represented, face high automation exposure. Without structured pathways into higher-value AI-linked roles, automation could stall or reverse gains in women’s workforce participation. AI applications in healthcare, agriculture, public transport, and financial inclusion can transform outcomes for millions of women if designed with their realities in mind.

From rhetoric to reform
As India convenes global stakeholders at the AI Summit, it has a rare opportunity to chart a distinctive model—one that integrates gender justice into technological leadership rather than treating it as an afterthought. Institutionalising this commitment will require more than side events and statements of intent. India could form a National Gender and AI Taskforce, which could serve as an anchor, embedding gender considerations systematically across AI policy, procurement, deployment and governance.

In the run-up to the Summit, the Indian Presidency hosted a panel on Gender-Responsive AI and Inclusive Digital Futures. The discussions underscored several starting points that could shape the mandate of such a task force:

·First, fix the data gaps. AI systems are only as representative as the data they are trained on. Tech leaders and public institutions must ensure the systematic collection of gender-disaggregated data across social, economic, and cultural domains. Without such data, AI systems built on incomplete data will inevitably reproduce that invisibility.

·Second, recognise that AI learns bias, not just facts. AI models are trained on historical records, employment-unemployment statistics, health, education, financial, social and economic records. If women are historically excluded, those are taken as “normal” rather than an anomaly. Correcting these inherited distortions requires conscious human oversight and corrective weighting rather than passive acceptance of historical baselines.

·Third, close the usage divide. A higher percentage of men, compared to women, are reportedly using generative AI, implying that AI ecosystems are male-dominated, even at the user level. India’s flagship digital initiatives, including Digital India, must explicitly address the gaps in AI adoptions through targeted outreach, digital literacy, subsidised access, and community-based training that enable greater female participation in AI usage and innovation.

·Fourth, institutionalise gendered algorithmic audits. Gender-responsive algorithmic impact assessments should become mandatory, particularly for AI systems used in hiring, credit scoring, welfare targeting, public service delivery and labour platforms. Audits must be designed not merely to detect bias but to correct it through iterative retraining and accountability mechanisms.

·Finally, place women at the centre of AI governance. Governance structures, including regulatory committees and standards-setting bodies, must ensure meaningful representation of women to ensure that the voices of those most affected by technological shifts are present at the table.

Moving from rhetoric to reform requires structural embedding. If India succeeds in doing so, it will not only build sovereign AI capabilities but will also build an AI ecosystem that reflects the democratic and developmental aspirations it seeks to champion. Therefore, with the India AI Summit, India has an opportunity to show the world that its technological future will not only be intelligent—it will also be equitable.

The authors are researchers at ICRIER. Views are personal.