India's IT ministry has issued an advisory requiring tech firms to seek approval before releasing AI tools, specifically those labelled 'unreliable' or under trial. The advisory, issued last Friday, emphasises that the availability of such tools, including generative AI, on the Indian internet requires explicit permission from the Indian government. This move aligns with global efforts to regulate AI, and India, a key market for social media companies, has been tightening regulations in this sector.

These are direct violations of Rule 3(1)(b) of Intermediary Rules (IT rules) of the IT act and violations of several provisions of the Criminal code. @GoogleAI @GoogleIndia @GoI_MeitY https://t.co/9Jk0flkamN

The advisory follows criticism of Google's Gemini AI tool by a top minister on 23 February, who accused Indian Prime Minister Narendra Modi of implementing policies characterised as 'fascist.' Google acknowledged the tool's occasional unreliability, particularly for current events and political topics. The advisory, which surprised many industry executives, also demands that tech firms appropriately label the fallibility of their AI model outputs. Some Indian startups and VCs express concerns that such regulation could hinder the nation's global competitiveness in the AI race.

Deputy IT minister Rajeev Chandrasekhar emphasised on social media that platforms have a legal obligation for safety and trust, highlighting that a mere apology for unreliability doesn't exempt them from the law.

The advisory also calls on platforms to ensure their AI tools do not threaten the integrity of India's upcoming general elections, expected to be held this summer, where the ruling Hindu nationalist party anticipates securing a clear majority.