Robots against Slavery?

Published on 18 October 2017

Pauline Oosterhoff

Research Fellow

Globally, an estimated 40.3 million adults and children are trapped in modern slavery. Slavery has many forms and does not automatically end with national economic growth. Much of the work that modern slaves do is repetitive, physically heavy and low-skilled manual work that can already be done by machines. One might hope that the modernization of production processes would eradicate modern slavery. Could automation and artificial intelligence liberate millions from the horrors of breaking stones by hand, embroidering handkerchiefs, picking and cleaning cotton, or painting plastic bangles? It could help to achieve the UN’s Sustainable Development Goal #8: “productive employment and decent work for all“.

Jobs at risk of automation

A substantial share of the workforce in all countries is at risk of automation. Rather than freeing slaves to pursue paid work, automation may simply deprive them of any work at all, and may put new pressures on other low-paid workers. Female workers in developing countries may be especially vulnerable to the impacts of automation. For example, in five ASEAN countries, more than 70 per cent of workers in the textile, clothing and footwear industries are women (Chang et al 2016). The co-existence of low-paid work and slavery in countries with rapid economic growth and technological development raises urgent questions about the relationship between automation, gender and modern slavery.

India is firmly committed to using digital technologies for development. Policies such as Digital India aim to deliver government services digitally and support the development of safe nationwide digital infrastructures. India also has the highest number of people living in modern slavery. They work in agriculture, in the garment and building industries (especially cotton mills, stone quarries and brick kilns), as well as in parts of the hospitality industry and domestic work.

High numbers of Tamil Nadu villagers in bonded labour

In Tamil Nadu, Southern India, tens of thousands of girls and young women have been recruited into employment schemes in the textile industry that result in excessive hours of work and extremely low pay – often with appalling effects on their mental and physical health. Poor conditions in the garment industry have received a lot of attention, and many improvements have been made. The spinning mills and power looms are still associated with higher levels of bonded labour than other parts of the garment supply chain. New IDS research in partnership with Praxis Institute for Participatory Practices, found that among 358 villages and hamlets near cotton mills in four districts in Tamil Nadu, only in a minority of the households (45%) were no family members working as bonded labourers. In 36% all of the family members were in bonded labour, and the rest had at least one. Most of these bonded labourers are girls and young women.

Girls earning well below the minimum wage

In the spinning mills, women and girls typically earn 180 rupees a day, well below the Indian minimum wage. Very high proportions of the mill workforce are given apprentice status, which allows employers to pay them lower rates (below 200 rupees per day). Many mill workers have to work six days a week, are not paid properly for overtime, and are sexually harassed. As a result their physical and mental health suffers.

These young women have an excellent understanding of the economic and gender dynamics of their situation. “We are in the factories because girls are seen as submissive and we are easier to control than men, both in the families and in the mill,” one explained. They remember a time when their fathers earned enough to support the family with irregular construction and agricultural work. Such opportunities have recently decreased because of climate change and other factors. But they are too young to remember that it was men who used to work in the mills. In response to successful unionisation in the 1990s, the textile industry started to target young women for employment. Bonded labour schemes such as Sumangali Thittam promised bonuses to pay for dowries, attracting women to the mills.

That allowed the mills to lay off their middle-aged male workers, many of whom took to drink, reinforcing their inability to find work. In 2002 alcohol sales were brought under the control of the state. Today revenue from liquor, and from the annual licences bars needed to sell it, are a major source of government revenue. Women’s and girls’ jobs at the mill help to pay for the men’s drinking habits.

Will jobs in the mills become redundant?

New developments in automation may soon make these female workers’ jobs in the mills redundant. Machines are more submissive and easier to control than young women. Yet where will these “liberated” women then go? Many of the girls have been forced to leave school early. Their earnings currently allow their siblings to stay in school. “Digital India” initiatives may ensure that these childrens’ mobile phones have better internet connections and a safe digital infrastructure. But will it help to create better jobs for them?

The Indian service industry has grown dramatically, creating millions of jobs in ICT, especially in the service sector. Studies have suggested that women may have an advantage in ICT, particularly in jobs such as helplines at call centres, due to superior social skills (Deming 2015). It is not for nothing that Apple and Amazon have chosen female voices for their automated personal assistant voices – they believe that people are more receptive to a female voice. But the existence of Siri shows that women in the IT service sector are vulnerable to replacement by AI, too. India’s services industry workforce is expected to shrink with a decline of 14% by 2021.

Even the oldest profession may be replaced by robots

Perhaps these young women will end up selling sex, voluntarily or involuntarily, possibly with the help of criminal trafficking networks. But even the oldest profession may ultimately be affected by automation: sex robots such as “frigid Farrah” are making rapid strides.  (Will such robots free women from the oppression of sex work? Or will they simply support and encourage male rape fantasies?)

The fact is that we do not have a sure-fire way to eradicate slavery in the existing economy. We are not really sure what role previous rounds of automation have played in either eliminating or encouraging modern-day slavery, and we do not know what effects new developments in automation and artificial intelligence will have. One thing we might want to do is to talk to young women working as bonded labourers in India’s cotton mills today, to bring their perspectives into the much needed debate on women and automation and the future of work. They may have a thing or two to tell us about how technology and automation could eradicate modern slavery, or exacerbate it.


The views expressed in this opinion piece are those of the author/s and do not necessarily reflect the views or policies of IDS.


About this opinion


Related content


Business Owners’ Perspectives on Running Khaja Ghars, Massage Parlours, Dance Bars, Hostess Bars, and Dohoris in Kathmandu, Nepal

CLARISSA Research and Evidence Paper 6

8 February 2024