Google apologises for Gemini results on PM, calls AI platform ‘unreliable’ | Company News

Google apologised to the Indian government for unsubstantiated comments made by its AI platform Gemini regarding Prime Minister Narendra Modi, revealed Union Minister of State for IT & Electronics Rajeev Chandrasekhar during an interview with the Times of India. The American tech giant described the platform as “unreliable” amidst growing concerns over the accuracy and integrity of AI-generated content.

Union minister Chandrasekhar highlighted India’s stance against serving as a testing ground for AI platforms, particularly amid global criticisms of disseminating biased, misinformed, or unverified information to users. He expressed dissatisfaction with AI data being released directly from labs onto the public internet without adequate testing or safeguards, leading to apologies when discrepancies are discovered.

According to the minister, Google responded to a notice from the Indian government regarding the unsubstantiated results produced by Gemini in response to a query about PM Modi by admitting that the platform is unreliable. Chandrasekhar criticised this response, stating that claiming unreliability is not a sufficient defence for AI platforms.

Google’s controversy with AI platform Gemini

Google’s Gemini, formerly known as Bard, introduced a text-to-image feature on February 1, which drew criticism on social media platforms like X (formerly Twitter). Users raised concerns about the platform generating erroneous results, including biased depictions favouring individuals of colour and inaccurate portrayals of historical figures such as the “Founding Fathers of America” and the “Pope”.

In India, Gemini sparked controversy for its response to a prompt about PM Modi’s political characterisation. The platform’s assertion that Modi has been accused of implementing fascist policies prompted further scrutiny. Meanwhile, responses to similar prompts about leaders of other nations, such as Ukrainian President Volodymyr Zelensky, were less contentious.

MeitY issues advisory for AI platforms in India

Amid these growing concerns about AI-generated content, the Ministry of Electronics and Information Technology (MeitY) issued an advisory on Saturday. The advisory mandates that all platforms ensure their computer resources do not permit bias, discrimination, or threats to electoral integrity through the use of AI or similar algorithms.

The advisory targets all AI models, large-language models (LLMs), and software using generative AI, requiring explicit permission from the government before deployment on the Indian internet.

While not legally binding, Minister Chandrasekhar stressed that compliance signals the future direction of regulation. He cautioned that failure to comply could result in future legislation imposing mandatory regulatory measures on platforms.

 

Originally appeared on: TheSpuzz