After stating multiple times that it will not regulate AI, the Indian Government last week released an 'advisory' on the deployment of AI products. This move sparked a wave of concern from the industry, leading to subsequent assurances on its intended scope from the Government.
The advisory can be accessed here; we have summarized the chain of events and current state-of-play here.
The Advisory: On March 1, 2024 the Indian Ministry of Electronics and Information Technology (MEITY) issued an advisory all Internet intermediaries and platforms that required them, among other matters, to obtain Government approval before launching AI products in the country.
This advisory was issued under the aegis of India's IT Rules, 2021, that prescribe due-diligence to be observed by online intermediaries who seek 'safe harbor' protection in India. All intermediaries were tasked with ensuring that any AI models, software, algorithms, etc., do not host or publish any unlawful content, or biases arising out of their AI models or platforms do not hamper the electoral process.
Para 2(c) of the 'Advisory' sparked the most concern, when it required that "the use of under-testing/unreliable Artificial Intelligence modes(s)/LLM/generative AI, software(s) or algorithm(s) and its availability to the users on Indian Internet must be done so with explicit permission..." of the Indian Government. A 'consent popup' mechanism was recommended to inform users about the fallibility of unreliability of the generated output.
The Government also asked intermediaries to ensure that any potentially misleading content is labelled with unique metadata or identifiers, allowing for the identification of its origin and the intermediary involved, in order to facilitate tracking of misinformation or deepfakes and their originators.Finally, the MEITY asked all intermediaries to submit an "action taken and status report" within 15 days of the advisory.
The Blowback: Given the subject matter, and how vaguely the Advisory was worded, not to mention its expansive scope, it is perhaps no surprise that it has resulted in a lot of questions and negative comment. It has been labelled a rushed foray into AI regulation, called out for being issued with ulterior motives, and criticized as putting the brakes on India's AI progress. The loose wordings used in the Advisory have not helped matters; for example, while it requires 'platforms' to comply with laws, this term is not defined in the parent 2021 rules. (It is also not clear what comprises an 'unreliable' AI model.) The requirement on all intermediaries to submit a 'progress report' to MEITY may be a non-starter too, with it not being clear how this reporting will be managed or enforced given the size of India's market.
The Clarifications: MEITY has since clarified that this advisory primarily applies to the large social media platforms and intermediaries, and does not apply to start-ups. It also limited the application of the Advisory to 'untested' AI platforms (though it does not clarify what is an 'untested' AI platform).
Where Things Stand: Post the backlash, the Indian Government seems to be downplaying the impact of the Advisory. It is also worth noting that the legal impact of an 'Advisory' is unclear; it is not a notification or direction issued under the parent law. Read in the context of India's upcoming national elections, this may be seen a (not very effective) attempt to police deepfakes and misinformation under the canopy of existing 'safe harbor' laws.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.