Google has extended an apology to the Indian government, labelling its AI platform, Gemini, as ‘unreliable’ following backlash over its biased response to a query about Prime Minister Narendra Modi.
What Happened: This development came to light after IT Minister Rajeev Chandrasekhar revealed the company’s reaction to a government notice that questioned Gemini’s controversial output. “They apologized, admitting the platform’s unreliability,” Chandrasekhar informed the Times of India, emphasizing that such an excuse was inadequate.
Chandrasekhar critiqued the direct transition of AI data from development to public use without adequate testing or safeguards. This practice, according to him, leads to situations where companies, caught off-guard, simply apologize for their platforms’ unreliability. He also highlighted that Gemini’s response violated Indian IT regulations and criminal codes, especially concerning bias and misinformation.
Further, the minister criticized AI platforms for offering consumer solutions while still in their trial phases, arguing against India being a testing ground for such technologies.
Increasing regulation: In a move to tighten regulation, the government declared on March 1 that all AI models undergoing testing or deemed unreliable must gain explicit government approval before being used in India. This directive aims to prevent AI-induced bias, discrimination, or any threat to electoral integrity.
In a related development, Google has paused Gemini’s feature of generating images of people after inaccuracies were reported in historical depictions. Alphabet CEO Sundar Pichai has acknowledged the platform’s offensive biases as ‘unacceptable,’ pledging structural adjustments to address these issues.
(The article is published under a mutual content partnership arrangement between The Free Press Journal and Benzinga)
