What is data feminism and its impact on AI

What does data feminism theory advocate?

The central idea is that data and algorithms are not impartial. As the majority of technology teams are made up of white men from rich countries, the systems end up reflecting only this worldview. The movement argues that choices about what data to collect and how to use it can unintentionally exclude or harm women, people of color and poorer people.

How did this trend reach big technology companies?

Big techs such as Google, Microsoft and Amazon have adopted the concept through image strategies and corporate practices. They started using terms like ‘responsible artificial intelligence’ and ‘data diversity’. This followed famous failures, such as facial recognition systems that were much more wrong in identifying black women than light-skinned men.

How is this theory influencing Brazilian legislation?

The concept has already reached the Federal Senate. The project that regulates artificial intelligence in Brazil requires companies to explain how their machines make decisions. The objective is to avoid discrimination based on race and gender, transferring responsibility for any errors in computer codes to human managers who control the system.

What are the main criticisms made of this movement?

Experts warn of the risk of data science turning into political activism. Critics argue that it lacks technical clarity: it is difficult to turn an abstract concept like ‘social justice’ into a mathematical rule that an engineer can test. Another fear is that the search for an ideological agenda ends up manipulating the data so that it only says what the authors want to hear.