Navigating AI Governance: Where Does the Balance Lie?

The recent advancements in Artificial Intelligence have created a new scope of services. Day by day the development in the AI services are improving leaps and the machines are becoming more powerful. This necessitates the governments to enact laws to govern the development, deployment, and use of AI.  EU has recently passed world’s first Artificial Intelligence Act to govern AI risks and practices, and USA has drafted an AI Bill of Rights. In India, the Ministry of Electronics and Information Technology (Meity) has issued an Advisory recommending practices for intermediaries and platforms providing AI related services.

Background

The advent of powerful AI models such as OpenAI’s ChatGPT and Google’s Gemini has garnered significant interest among the public in artificial intelligence’s potential applications and ethical implications. As we all know, these large language models (LLMs) can create original content such as text, images or computer program code using large sets of data.

In recent times, it has stirred up the issue of inventorship and ownership of such original content, raising complex questions around who should be credited as the inventor or owner of such original content generated by AI models. The common question that arises here is, “Who is the true owner or inventor?” Also, it is more pertinent to note that, to receive any such content from these AI models, a prompt from the user is required. In that case, other questions that arises are, “Will the user providing the prompt have, or claim ownership or co-ownership?” or “Will it be considered public domain, as anyone with the same prompt may get a similar output?”

EU and USA Paving the Way with laws governing AI:

Existing IP laws that intend to protect the owner of the original or novel work do not squarely apply to the outputs generated by these AI models. We cannot deny the fact that improving machine capabilities, in light of AI developments, is helping humans and making machines more powerful than ever. This unique set of factors necessitates a comprehensive legislation to govern the ethical practices and risks arising from use of AI and Machine Learning.

Earlier this year on the 13^th of March 2024, the European Union passed the world’s first Artificial Intelligence Act, and it is based on the risks of AI (Risks which are unacceptable, high, limited, and minimal) and one category for general-purpose AI.[1] Further, the federal government of the United States has also drafted legislation on the subject matter titled “AI Bill of Rights.”

Position in India: Non-Binding Advisory

While India has no specific legislation governing AI, the Government of India published an AI Advisory that recommends certain practices be adopted by intermediaries and platforms that provide AI services, and their users.

The said advisory was issued by the Cyber Law and Data Governance Group of the Ministry of Electronics and Information Technology (Meity) on the 1^st of March 2024 with the subject “Due diligence by Intermediaries / Platforms under the Information Technology Act, 2000 and Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021[2]” raising concerns regarding the failure of intermediaries and platforms providing AI services to undertake due-diligence obligations imposed on them under the IT Rules (Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021).

Even though the advisory is not binding in nature, this advisory requires intermediaries and platforms to adopt certain practices if they are developing and making available their services, such as Artificial Intelligence model(s) /LLM/Generative AI, software(s) or algorithm(s) (“AI services”), for use by the public.

We can classify this Advisory into three prongs with respect to its applicability viz., all the intermediaries and platforms providing their AI services and the users of such services. A summary of the relevant advisory is provided below:

1. Intermediaries and platforms that provide AI services. It is advised to be complied with, in addition to the advisory 2(4)/2023-CyberLaws December 26, 2023. This advisory has four specific advice that the intermediaries and platforms should comply with:-
   1. Compel users NOT to deal with unlawful content while using the AI services;
   2. NOT permit bias or discrimination or threaten the integrity of electoral processes;
   3. Necessary to have consent pop-up informing the users about the fallibility or unreliability of the output generated based on their undertested or unreliable AI models;
   4. Inform users through Terms of Service and User Agreements on the consequence of dealing with unlawful information.

2. Creator Label / Identifiers: It advises the intermediaries and platforms, as the case may be, to use permanent labels or identifiers on their synthetic outputs, such as text, audio, visual, or audio-visual information, to identify the creator. It is informed by the Ministry that this advice is provided to curb misinformation and deepfakes and to find the creator or the first originator of such misinformation or deepfakes. (Advisory 3)
3. Violations & Penalties: The final advisory reiterates that non-compliance by any users with the IT Act or IT rules will result in a potential consequence for intermediaries, platforms, and even users who violate such laws and other criminal codes. (Advisory 4)

Way Forward

Given the fact that India is a leader in software development and Software as a Service (SaaS) and is one of the largest consumer of AI services with its swelling population and penetration of smartphones and internet connectivity, it is crucial and about time for India to have a stand-alone substantive legislation governing AI practices.

[1] European Parliament legislative resolution of 13 March 2024 on the proposal for a regulation of the European Parliament and of the Council on laying down harmonised rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union Legislative Acts (COM(2021)0206 – C9-0146/2021 – 2021/0106(COD))

[2] https://www.meity.gov.in/writereaddata/files/Advisory%2015March%202024.pdf