Chennai: The Centre for Responsible AI at IIT Madras launched IndiCASA on Tuesday, a dataset that will be useful for bias risk detection and assessment in language models within the Indian context.