It is such a pleasure to write these notes during the big festive season in India. Last week was Dussehra which celebrates the triumph of good over evil in many forms. In other parts of the country, it was Durga Puja, which celebrates Shakti or female power of the universe. Whatever the festival, it is time for great food, family and friends, getting together and plenty of shopping, with big sales happening everywhere.
Amidst this, the 8th World Telecommunication Standardization Assembly and the India Mobile Congress were inaugurated 15 Oct. by Prime Minister Narendra Modi. It is the first time that India and the Asia-Pacific region hosted the WTSA.
The prime minister's speech gave an indication of the policy and regulatory direction in the digital governance domain that India is headed in. He called for the creation of "global standards for ethical AI and data privacy that also respect the diversity of different countries" and asked members of the WTSA to "think about how to make telecommunications safe for everyone."
"In this interconnected world, security cannot be an after-thought. India's Data Protection Act and National Cyber Security Strategy show our commitment towards creating a safe digital ecosystem," he said. "I also want to put forward this issue before a global platform like WTSA. This issue is of global framework of digital technology, global guidelines. Now the time has come that global institutions will have to accept its importance for global governance. Do's and don'ts will have to be made at the global level for technology."
The focus on creating global frameworks that are inclusive and accommodate for diversity is a key takeaway. Interestingly, at the same event, leaders from the Indian telecom industry pushed for reforms in policy to ensure data localization, among other things.
Ten days earlier, 5 Oct., Minister for Electronics and Information Technology Ashwini Vaishnaw talked about India's plans for artificial intelligence regulation. Speaking at the Economic Times Startup Awards, he said, "We have two different models before us. One is the US model, second is the European model. I honestly don’t think that the way some of the countries have put so much regulation on AI, I don't think that is the right way to go.
"[T]he way we created a structure in our data privacy law and in the telecom law, where we allowed the space for innovation to flourish, that space we gave while we made sure that the harms which can come to the society are fully controlled. That approach is the approach we think we will be able to take in the case of AI also," he said.
Food for thought?
Meanwhile, speculations on what is expected in the Digital Personal Data Protection Act Rules keeps coming up in the media. On 19 Sept., the Economic Times reported the rules are expected to prescribe umbrella guidelines and frameworks on consent management rather than exact specific rules.
Other indications of what is in store can be gleaned from a few other developments around us.
For example, we have what the Parliamentary Standing Committee on Communication and Information Technology will take up in 2024-25. According to a bulletin from the Lok Sabha published 8 Oct., the committee is slated to look at the impact of AI and related issues, social and digital platforms and their regulation, the emergence of new forms of currencies and their impact, and the regulation and monitoring of digital and cybercrimes.
And then there is the pushback on the DPDPA from the All India Game Developers Forum. In a report the forum recently released, they outlined the challenges involved in handling the requirements pertaining to children, given many of their users are below the age of 18. Specifically, obtaining parental consent as well as not conducting behavioral monitoring or targeted advertising to children are seen as a "disproportionate challenge."
It will be interesting to see how these play out in the coming months and years.
Meanwhile, developments in the AI domain in India are in what I call an "accelerated" mode.
National eGovernance Division CEO Nand Kumarum recently spoke about the IndiaAI Datasets Platform being live by January 2025. This platform is one of the pillars of the government's INR10,000 crore IndiaAI mission. Data from central and state governments as well as from the private sector will be a part of this platform.
"The idea primarily is like HuggingFace — you have models, you have datasets, and you have people coming up and using those datasets and building models. We are trying to do something similar," Kumarum said.
Meanwhile, an interesting development has been the launch of BharatGen by the Ministry of Science and Technology 30 Sept. Funded by the government, the generative AI initiative is focused on creating multimodal LLMs for Indian languages. BharatGen and its applications would be available as a public good with four key distinguishing features: multilingual and multimodal nature of foundation models, "Bhartiya" dataset-based building and training, open-source platform and development of a generative AI research ecosystem in India.
Another interesting development — this one from the private sector — has been the announcement by Tech Mahindra of its LLM trained in Indic languages titled "Project Indus." Initially trained in Hindi and 37 dialects, the company said it would cover other languages and dialects over time.
Availability of LLMs trained in Indian languages is a huge deal, given how much data is in India's many local languages.
Amidst all this hoopla, we also heard some words of caution from the Reserve Bank of India, India's central bank. At an event in New Delhi, RBI Governor Shaktikanta Das said, "The heavy reliance on AI can lead to concentration risks, especially when a small number of technology providers dominate the market." He added that this could exacerbate systemic risks as failures or disruptions in these systems could spill over into the larger financial sector. He also warned that AI's “opacity” makes it tough to examine algorithms which help lenders make decisions and could potentially lead to "unpredictable consequences in the market."
This gives us an indication of what actions the RBI may take down the line in the sphere of AI. The RBI has been one of the most pro-active regulators in the country while addressing digital risks.
Meanwhile, the data juggernaut continues to march on. The latest report from the Telecom Regulatory Authority of India says the total number of internet subscribers in India increased from 954.4 million at the end of March 2024, to 969.6 million at the end of June 2024. Out of 969.6 million internet subscribers, there are 42.04 million wired Internet subscribers and 927.56 wireless Internet subscribers.
As we all in India look forward to Diwali in a couple weeks — with the firecrackers that typically come along with it — I wonder what other "crackers" are around the corner for us in the digital space, as well.
Shivangi Nadkarni is co-founder of Arrka, a Persistent Company.
