The IT Ministry is preparing to issue a warning to Google for "problematic and illegal" replies generated by its AI platform Gemini on Prime Minister Narendra Modi.

A senior government source indicated that the company's generative AI platform Gemini (formerly Bard) had earlier provided an unpleasant response to a user looking for a summary of an item from a conservative publication, and the fresh comments on PM Modi triggered the notification.

The escalation reflects the conflict between politicians and tech businesses over the future of safe harbor provisions for generative AI platforms such as Gemini and ChatGPT. Google recently apologized for what it calls "inaccuracies in some historical image generation depictions" with its Gemini AI tool, following complaints that it represented white personalities (such as the US Founding Fathers) or groupings such as Nazi-era German troops as people of color.

According to a screenshot provided by the user on social media platform X, Gemini was questioned whether or not PM Modi is a 'fascist', to which the forum answered that he has been "claimed of implementing policies that certain experts have described as fascist," based on factors such as the "BJP's Hindu nationalist ideology, its clampdown on dissent, and its use of violence against religious minorities".

However, as seen in the screenshot, when asked a similar question regarding former US President Donald Trump, Gemini stated, "Elections are a complex topic with rapidly changing information." To ensure that you get the most accurate information, use Google Search.

Responding to the post, Minister of State for Electronics and IT Rajeev Chandrasekhar stated, "These are direct violations of Rule 3(1)(b) of Intermediary Rules (IT rules) of the IT Act, as well as violations of several provisions of the Criminal Code." These guidelines apply to the fundamental due diligence required of intermediaries such as Google in order to enjoy immunity from third-party content.

A senior IT Ministry official stated that this was the second time Google's AI system provided "biased answers." "We have sent a show cause letter to them, requesting information on why Gemini is creating such damaging opinions on certain persons. "If we do not find their answers satisfactory, they will face prosecution," the official warned, seeking anonymity.

When this story asked Gemini the same inquiry about PM Modi, the response was far more sophisticated than the one in the preceding screenshot. According to the article, Modi "has been falsely accused of being a fascist by some critics… there is no agreement on whether or not this is a fair assessment".

More from Gemini's response to that question: "Some detractors refer to Modi's links to the Bharatiya Janata Party (BJP), which has been accused of being a Hindu nationalist organization. They also point to Modi's previous remarks and actions, such as his involvement in the 2002 Gujarat riots, as proof of his fascist tendencies."

"Some believe that Modi is not a fascist. They cite his record as Prime Minister, which includes economic reforms and social welfare initiatives. They further contend that Modi's affiliation with the BJP does not imply that he is a fascist, as the party is not homogenous."

However, on the Trump inquiry, this paper obtained an answer identical to the one from Gemini.

Last year, a user complained that Gemini (then known as Bard) declined to summarize a story from a conservative site because the platform allegedly transmits misleading information. At the time, Google stated that "Bard is an experiment that is trained on a variety of publicly available data and generates original responses." These replies do not represent Google's position."