Avirup Bose (Professor of Competition Law and Policy, JGLS)
Leela Tarang Krishna Paladugu (Assistant Professor, JGLS)
Mr Parvesh Kumar Sharma
Does the proposed US Algorithmic Accountability Act serve as a blueprint for a "sector-neutral" framework, to govern all the AI or algorithmic processes across the sectors? In short, is this applicable for India!
India has developed a DPDP Act (Digital Personal Data Protection Act). The Act is India’s most consequential imprint on Artificial Intelligence. Ethics, data handling and a score of other issues are covered by the DPDP Act. Commentaries by well-known firms say that the DPDP Act has a far reaching impact. The US didn't have a separate data regulation to match so far.
However, while India has the DPDP Act, it lags the USA in integrating AI into each and every aspect of business. India’s AI market remains largely in the experimental stage. So, we are still struggling with a lot of copyright issues, intellectual property infringement etc., as opposed to the US, where they have already integrated AI deep into their military and global and political operations.
AI technology in India is still in its nascent stages of integration. So having a regulation of this massive scale may be seen as taking the step too early. In fact, there is now the concern that with the DPDP we have a rival to the CCI (Competition Commission of India), who, although they are absolutely different entities, govern how players are collecting data, and their usage. No wonder there is already a lot of debate about a significant amount of overlap between DPDP and CCI about policing of operational practices by businesses using the data that they have collected.
Apart from this overlap, there is also the Digital Competition Bill coming in, which seeks to restrict even more what companies can and cannot do - thus compounding the regulatory clutter. So, there is too much chaos happening in the market. Therefore, it is still debatable whether India requires a separate AI regulation on top of these. Since there is still so much to be discovered and uncovered in the AI field it is preferable that India is taking a decentralized approach with individual sector experts making their own rules for AI within their domains.
In contrast there is one bill in the US Senate called CREATE AI Act, which is scheduled to reach bipartisan consensus and is almost scheduled to be the first AI bill that may pass the Senate (as reported on 1st of July, 2026). It is not known however, whether it will pass the Congress.
Yet in the Indian context, just a few days back the Member Secretary of the Ministry of Electronics and Information Technology (MeitY) has publicly said that they are already in the process of developing a dedicated AI bill. Mr Ashwini Vaishnaw (Minister for Railways, Information & Broadcasting and Electronics & Information Technology) has also said that the time has come for India to have its own AI regulatory framework. He did not use the term bill, but the Secretary used the term bill, and has said that the current framework is not working. He also clarified however that MeitY only drafts the bill and cannot control when parliament passes it.
Now, the Digital India Act has now been abandoned. Thus, these very recent developments of both the Minister and the Secretary speaking on similar lines (within a span of a few weeks) definitely suggest lots of movements happening towards creation of a dedicated, sector-neutral AI bill. It seems that we are no longer having that zone where India is speaking about what it was very famous for - the light touch approach or the innovation friendly approach. The Mythos saga has possibly made India realise that it is not at the regulatory table and that has prompted a shift toward active policy-making.
In effect, India seems to be publicly acknowledging at the ministerial level that they are in the process of drafting a sector-neutral kind of bill.
Background:
There were 2 articles on AI that were published
in a prominent financial newspaper and a web portal.
The first article titled “Indian banks may face an AI compliance bill: Who will bear the cost?” - is a continuation of the CRG team discussions that's happening from last week with regards to how RBI came forward with the AI compliance bill where banks and financial institutions are required to mandatorily implement kill switches. They are required to have a board that will oversee all these particular AI aspects. One of the biggest stakeholders in the entire RBI sector are the financial institutions and the concern has been how it's going to increase costs for people, banks, financial institutions, and everyone that are involved in there.
The article talks about how a lot of it is going to bring in a new form of business in terms of AI based insurance or AI based analysis of portfolios of people. And it also speaks about how RBI is increasing the cost of governance with regards to implementing AI. So, it's a bit of a mixed approach as to what the ultimate effect of this particular governance could be, in terms of what RBI is planning to do.
In short, analysis needs to be done whether it is actually increasing the cost of compliance and cost of governance for these entities. It also talks about several businesses coming into picture, but unless there is actual proof of that in terms of AI insurance, and also with regards to analysing the data models that are present, and how RBI reacts to that, there is still a lot of work to be done in seeing how this plays out in the real world.
The second article titled “AI isn't the biggest regulatory worry for Indian startups, data governance is: Survey” is a survey that was done by Mint, and it talks about how, for startups, AI is not the biggest concern, but rather the data governance aspect is.
The survey indicates that Indian startups have been adversely affected as there are lots of regulations pertaining to data governance that they have to fulfil. It also talks about how investments in businesses and in AI will go down because of that. In short, there are lots of regulatory issues for the new and the existing Companies.
A few weeks back, there was a discussion among CRG panellists regarding why investments in AI are going down and where India is not investing in its own AI. These two events have highlighted the massive hurdles being created by strict regulatory requirements.
This study explores a persistent data governance issue happening for quite some time with regard to fear within the private sector - struggling with regards to building AI or investing in AI. Specifically, it investigates whether there is actually a new struggle or whether it's a continuation of existing challenges. Data regulation and data governance is not a very new thing, or it's not like India has developed new regulations right away the moment AI came in. RBI came up with the compliance bill, but it's a little vague as to how the compliance bill will actually drastically or negatively impact the stakeholders there. It is a pre-emptive action, making sure that the AI in banking and finance is not too uncontrolled.
Sebi has recently launched its own high-powered committee on AI in the securities market, while RBI has its own AI compliance focus. Maybe these 2 issues could be combined into a broader topic as how regulators are kind of looking into this aspect of AI regulation from the point of view of the current anthropic saga that has happened.
Leave a Reply