We need to focus more on building India’s AI ecosystem: MeitY Additional Secy Abhishek Singh

Copyright indiatimes

ET OnlineSanjeev Bhikchandani, Founder and Executive Vice Chairman, Info Edge and Abhishek Singh, Additional Secretary, Ministry of Electronics and Information Technology at TiEcon Delhi NCR 2025. Artificial intelligence (ML) has quickly taken over the world, leaving no sector untouched. India, known for its top technology talent, needs to focus more on building its own AI models that differentiate themselves from those in the West, said Abhishek Singh, Additional Secretary, Ministry of Electronics and Information Technology.He said that India’s strength has been software, and Indian-origin people are part of almost every global big tech company.“India has the potential of becoming a use case capital for AI. There is a lot of talent around. There is a lot of support from the government—whether in the form of computing, building foundation models, or providing data sets. We are working towards getting Indian companies to build their foundation models, ensuring that Indians use Indian models today,” he said at a session titled ‘How India Can Win the Global AI Race’ at TiEcon Delhi NCR 2025.He noted that foreign AI companies such as OpenAI made ChatGPT freely available for all Indians to encourage the use of their AI models. OpenAI, led by Sam Altman, has recently announced that its mid-tier plan ChatGPT Go, which normally costs Rs 399 per month, will be available for free of charge in India for one year starting November 4, 2025.“But in this world, nothing is free. Whenever a product is free, then you are the product. The same thing applies there. So, all these models, all these companies who have made models, Google or Grok or Meta, they know that Indian companies can come out with their own models. They will want to ensure that more and more Indians start using their models so that the data from Indians start coming to them so that their models become much better,” he said.Live EventsIndia is actually playing a catch-up game, he said, expressing hope that Indian companies will come up with the Indian model so that the country’s IT, knowledge, intelligence, and data sets stay within Indian hands.He acknowledged that India, unlike China which restricted the presence of Google or Meta, has traditionally been open to Western companies operating in the country. “If we have to do that, maybe it will be regarded as a regressive step. It will be regarded as what China did. There are larger issues involved with regard to whether we restrict the operations of big tech companies within India and allow only Indian companies.”“India should benefit from Western AI too,” he said, “as they are customising their applications and products for the Indian market, improving sectors like agriculture and healthcare.” However, at the same time, it is important not to lose focus on building up the homegrown ecosystem, he added.“So very often some companies, even within the government, when we use AI, we know that we cannot let government data go through servers outside India. So, what we do is that on our own servers, we download the open-source models, whether it’s Llama or something else, and then we build applications on that so that we retain the data within our premises. So similarly, Indian companies in the corporation will have to think of doing something similar,” he said.Talking about the impact of AI and code generators on India’s IT sector and talent, Singh emphasised that Indian companies like Tata Consultancy Services (TCS) and Infosys should build their own code generators to retain the expertise of writing efficient code within our own teams.“What is happening with the reliance on code generators like Cursor AI or Copilot is that many developers across the whole ecosystem have been using these tools, and their ability to write code is being transferred to these models. In India, out of the 6 million people who are employed in this sector, many of them do low-value coding or debugging. So, if these tools are trained on doing this, they can start doing the same job better, and then many people in India will have the risk of losing their jobs. So, what is required is that Indian companies, whether it’s a TCS or Infosys, with their huge repository of code available with them, should actually be building their own code generators to retain the expertise of writing efficient code within our own teams, within our own geographies, within our own company,” he said.All Indian students studying coding in various colleges and those in entry-level jobs should start using code generators, he said. “This will require a massive effort.” He mentioned that the government has been talking to several companies and organisations, like NASSCOM, about how to enable this.In July this year, the government launched the Skilling for AI Readiness programme targeting school students from class 6-2 and educators to focus on topics like machine learning (ML) and the ethical use of AI. The Union Budget 2025-26 also allocated Rs 500 crore to establish a Centre of Excellence in Artificial Intelligence for education, aimed at advancing AI-driven learning and innovation.Add as a Reliable and Trusted News Source Add Now! Read More News oninfosysIndia AI ecosystemMeitYABHISHEK singh meityChatGPT IndiaAI talent IndiaAI education IndiaSkilling for AI Readinesstata consultancy services 20 cities. Countless breakthroughs. Secure your spot at the ET Make in India SME Regional Summit.....moreless Read More News oninfosysIndia AI ecosystemMeitYABHISHEK singh meityChatGPT IndiaAI talent IndiaAI education IndiaSkilling for AI Readinesstata consultancy services20 cities. Countless breakthroughs. Secure your spot at the ET Make in India SME Regional Summit.....moreless Prime ExclusivesInvestment IdeasStock Report PlusePaperWealth Edition123View all Stories