MEITY Advisory for Intermediaries on Use of Artificial Intelligence Models

Among the first attempts to regulate the artificial intelligence landscape of India, the Ministry of Electronic and Information Technology (MeitY) published an advisory on March 1, 2024, notifying intermediaries/platforms under the Information Technology Act, 2000 and Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (“Intermediary Rules”) on the use of Artificial Intelligence model(s)/LLM/Generative AI, software(s) or algorithm(s) (collectively “Models”) by the intermediaries on or through its computer resource. The key provisions are produced hereafter.

  • The intermediaries to ensure that the use of Models does not permit the users to host, display, upload, modify, publish, transmit, store, update or share any unlawful content as outlined in the Information Technology Act and Rules thereunder.
  • Intermediaries to seek explicit permission from the Indian Government prior to usage and deployment of under-testing and unreliable Models. Where such under-tested and unreliable Models are used, such Model has to be: (a) appropriately labelled to ensure that the users are aware of the inherent fallibility and unreliability of such model; and (b) a “consent pop-up” mechanism has to be used which would explicitly inform users about the possible and inherent fallibility or unreliability of the output generated.
  • Intermediaries providing software and/or other computer resource which facilitates synthetic creation, modification, or generation of information and may be used for generation of misinformation or deepfake content are advised to embed a permanent unique metadata or identifiers into such created content. These identifiers shall assist in tracing back to the intermediary, the user of the software and/or resources and/or first originator of such misinformation or deepfake.
  • All intermediaries shall ensure that their computer resource do not permit any bias or discrimination or threaten the integrity of the electoral process including via the use of the Models.