Government to IndiaAI Mission LLMs: Fix bias in AI models

The Government of India has directed large language model (LLM) developers backed by IndiaAI Mission to fix biases in AI models, reports Economic Times. Official at the ministry of electronics and information technology (MeitY) told the publication that given the country’s unique diversity, and society, LLM developers must ensure that government-supported foundational AI models should not throw up insensitive results when confronted with challenging prompts, officials at the.

AI models may get infused with bias and disparities if the data they are trained on are embedded with historical biases or systemic disparities embedded. Bias mitigation is the process of systematically identifying and reducing unfair prejudices in these LLM models.