The Ministry of Electronics and Information Technology (MeitY) has issued a new advisory that requires digital platforms to seek prior approval from the government before launching any artificial intelligence (AI) product or service in India. The advisory also mandates platforms to label under-trial or unreliable AI models and ensure that they do not host any unlawful or biased content.
The advisory, which was issued on March 1, 2024, is a first-of-its-kind globally and aims to regulate the use and development of AI in the country. Further, the advisory by the govt. states that all AI models, large-language models (LLMs), software using generative AI or any algorithms that are currently being tested, are in the beta stage of development or are unreliable in any form must seek “explicit permission of the government of India” before being deployed for users in India.
In addition, the advisory also directs platforms to ensure that “their computer resources do not permit any bias or discrimination or threaten the integrity of the electoral process” by the use of AI, generative AI, LLMs or any such other algorithm. The platforms are also required to offer their services to Indian users only after “appropriately labelling the possible and inherent fallibility or unreliability of the output generated”.
The advisory warns of penal consequences for non-compliance with the directive and asks platforms to submit a report on the status of their AI products and services within 15 days of the issuance of the advisory.
The advisory comes in the backdrop of a recent controversy over Google’s AI chatbot Gemini, which was accused of being biased and malicious when asked questions about Prime Minister Narendra Modi and other prominent global leaders. Google later said it was working to fix the issues and was temporarily stopping Gemini from generating images as well.
The advisory also follows the statement of Union Minister of State for Electronics and Information Technology Rajeev Chandrasekhar, who said that draft laws to regulate the AI space would be issued by June or July this year.