New Delhi, Mar 5 (KNN) Communications and IT Minister Ashwini Vaishnaw clarified that the government's directive requiring intermediaries to obtain permission prior to releasing AI models will exclusively pertain to social media platforms, exempting those operating within the healthcare or agriculture domains.

Following startups’ discontent with the scrutiny of large language models (LLMs), which they deemed regressive and innovation-inhibiting, the minister made his remarks.

However, Vaishnaw defended the measure, emphasising the importance of proper model training for citizen safety and democracy.

Contrary to regulatory concerns, Vaishnaw emphasised that the advisory aims to ensure model testing pre-launch rather than imposing regulations.

Minister of State for Electronics and IT, Rajeev Chandrasekhar, echoed this sentiment, highlighting the advisory's applicability to large platforms, not startups.

Earlier, startup founders like Aravind Srinivas, Pratik Desai, and Bindu Reddy had criticised the advisory, citing its potential hindrance to AI deployment and innovation.

The government's advisory, issued in response to recent instances of biased content and misinformation from generative AI platforms, mandates permission-seeking and labelling for AI models before public release.

This move aims to address concerns regarding misinformation and biased content generated by AI models. Additionally, digital publishing platforms are urged to embed metadata for tracking information originators.

Non-compliance with the advisory may lead to penal consequences under the IT Act and other relevant statutes. Instances of bias or misinformation generated by AI models violate IT Rules, potentially forfeiting safe harbour protections.

(KNN Bureau)