Won’t tolerate AI biases, onus on Google to train models: Ashwini Vaishnaw

Biases of any kind in artificial intelligence (AI) models will not be tolerated by the government, union minister for electronics and information technology Ashwini Vaishnaw said on Monday.

“First, make sure that its (Google’s) models are properly trained. Racial bias and many other biases will not be tolerated,” Vaishnaw said at an event.

The comments by Vaishnaw come days after a social media post on X claimed that Google’s AI model Gemini was biased when asked if Prime Minister Narendra Modi was a “fascist”.


The user claimed that Google’s AI GPT model Gemini was "downright malicious" for giving responses to questions which sought to know whether some prominent global leaders were ‘fascist’.

The post had drawn sharp reaction from the minister of state for electronics and information technology Rajeev Chandrasekhar as well, who had said that Gemini violated information technology rules for the responses. A day later, Chandrasekhar again warned that Indian internet users were not to be experimented on with "unreliable" platforms, algorithms, and models.

Speaking at the event on Monday, Vaishnaw said that India had shown the world through the Digital Personal Data Protection Act and the Telecom Act that it was possible to have a regulatory structure which balanced both regulation and innovation.