Ministry of Electronics and Information Technology updates its advisory on the deployment of AI by internet companies in India

The advise on the use of artificial intelligence (AI) by internet businesses in India has been amended by the Ministry of Electronics and Information Technology (MeitY). Companies had to first get government clearance before releasing AI models that were considered "under-tested" or "unreliable." The most recent advise, which was released on March 15, removed the mandate and replaced the earlier March 1 order. 

The businesses are still obliged to adhere to the new rules right away though the procedure has been streamlined. The updated standards mandate that AI models, such as generative AI software and large language models (LLMs), be tagged with a warning about the possible unreliability of their outputs. 

The purpose of the action is to educate consumers about the fallibility of material produced by AI. Furthermore, permission pop-ups must be implemented by the platforms to warn users that they could come across inaccurate or untrustworthy information. This strategy aims to strike a balance between innovation and user awareness and safety.

The revised advice also stresses how important it is for AI models to refrain from disseminating anything that is illegal in India, including anything that may be interpreted as biassed, discriminatory, or endanger the democratic process. 

Additionally, the government has directed platforms to mark or embed material with unique identifiers in order to detect deepfakes and falsehoods. The goal of the measure is to make it simpler to identify the original source of material that has been altered or artificially manufactured.