Artificial Intelligence can make biased decisions based on societal prejudices: CJI Chandrachud

CJI Chandrachud
File Photo: CJI DY Chandrachud

July 24, 2023

Chief Justice of India D Y Chandrachud recently said many Artificial Intelligence (AI) systems have been shown to exhibit biased decision-making based on data inputs that reflect societal prejudices. He further said AI, which has immense potential, can also be used to perpetuate discrimination and unfair treatment.

“This is because technology doesn’t develop in a social vacuum. They take place inside an intricate tapestry of societal realities and legal, economic and political structures” he said.

“A significant impact of AI is its potential to amplify discrimination and undermine the right to fair treatment. Many AI systems have been shown to exhibit biased decision-making based on data inputs that reflect societal prejudices. For example, AI recruitment tools deployed by firms favoured men over women because the tools were trained on the profile of successful employees who, for gendered reasons, happened to be predominantly male. In this, data-driven systems can perpetuate biases and marginalise the social control mechanisms that govern human behaviour,” Chandrachud said while speaking at the 60th Convocation Ceremony of the Indian Institute of Technology (IIT) Madras.

“Today Artificial Intelligence are the two words on everybody’s lips. AI is being deployed rapidly to enhance the efficiency of computer software and hardware. In a relatively short time, AI tools such as ChatGPT has shown that even computer can crack a joke, write codes or even draft legal memos. Even in the Supreme Court, we are using AI on a pilot basis for live transcriptions of court proceedings. The trepidations voiced by many about the advent of AI is an example of how new technologies do not exist in a vacuum. Always keep in mind what values AI represents and what affordances it creates. I have personally been fascinated with the AI image-generating software, which allows you to put in a written prompt and generate an image, but I am also cognizant of its potential misuse to create deep fakes, to mislead, and to bully individuals,” the CJI added.

Share with others