The Perception of Technology and Data, and its Impact on Data Governance Decisions

A study of women farmers in agricultural co-operatives in India

Malhotra Himmat

Master’s Thesis, 2022. Advisor: Patrick Shulist

AI has huge potential in farming, particularly in farmer cooperatives, but before data governance frameworks can be developed, it’s crucial to understand how communities perceive data and technology. In low-literate communities, socio-cultural norms can significantly affect data and technology perception, which, in turn, influences data governance decisions.

Based on a study of the Megha Mandli agricultural co-operative, women farmers accept technology initiatives that align with their community values and beliefs. Empowerment, pride, status, and community welfare are significant values for these women. Hence, technology that allows them to train themselves, improve their farming produce, gain better access to markets, verify information, and share their work with the world is well received – as it evokes a sense of empowerment and pride.

In low-income settings, there is often a wide range of data literacy and access, leading to a hybrid model of data sharing and governance within cooperatives. This model relies mostly on the values of trust and sisterhood among members, with those with good access to data technologies acting as ‘conduits’ of data insights to those without access. While this mechanism is contextually relevant and in line with community norms, it can prove problematic in the future, as there are no formal systems in place to hold conduits accountable for any misuse or breaches of data.

Formalising the roles and responsibilities of these members and capacity building to bridge gradients within cooperatives are crucial steps. It can help mitigate the risk of data breaches and misuse, as well as avoid a situation where power is accumulated within the hands of the conduits.

While trust within the co-operative communities is an important asset and enabler of data sharing, it can make a very essential part of data governance redundant – that of obtaining ‘Informed Consent’. Usually, members are likely to give their consent not through understanding and evaluating what is being asked of them, but merely based on the trust they hold for the co-operative. As such, co-operatives should implement a ‘granular consent’ mechanism on data sharing and use, breaking down the flow of data each time it needs to be shared in order to obtain consent for that particular action. This would make it easier for women farmers to understand where their data is going and how it’s being used.

In summary, understanding community values and beliefs, formalising data governance roles, and implementing a ‘granular consent’ mechanism are crucial steps for participatory governance and effective AI implementation in farmer cooperatives.

 Read the full thesis