India Developing Its Own AI Chip, Plans Generative AI Model With 18,693 GPUs

India is developing its own AI chip and a generative AI model to reduce dependence on foreign technology: reports ETV Bharat's Surabhi Gupta.

New Delhi: India has reached considerable milestones in the realm of artificial intelligence (AI), placing a new emphasis on the local development of AI chips and large language models (LLMs) to decrease reliance on foreign technology and strengthen its stance in the global AI playing field.

The Ministry of Electronics and Information Technology (MeitY) in cooperation with the Centre for Development of Advanced Computing (C-DAC) has started designing India's own capability in AI chip. The effort, which aims to develop hardware that can run generative AI models like ChatGPT efficiently, is expected to be completed by 2027. The action aims existing worries about diminished access to advanced processors from firms like NVIDIA, which are necessary for teaching and running advanced AI models.

Generative AI Models and Cultural Relevance

In a statement made during the Utkarsh Odisha Conclave on January 30, Union IT Minister Ashwini Vaishnaw announced India commencing work on its own generative AI model, following in the footsteps of global players such as OpenAI's ChatGPT and China's DeepSeek. He also asserted that the India AI Compute Facility would support the initiative with an impressive stockpile of 18,693 GPUs to spur the development of an LLM catered to India.

India has set to work to build a foundational AI model that would take cognizance of conventional arts and culture diversity in the country to eliminate biases from the datasets. Proposals for the AI model development will be invited, as per Vaishnaw, who added it would take 4-8 months for its development.

Besides hardware development, India is working towards funding and deploying indigenous LLMs. These models can both understand and generate text with a human level of fluency and provide applications across the domains of healthcare, education, and customer service. India wants to develop its own LLMs to address its linguistic and cultural uniqueness with the view of making AI more context-relevant and effective.

The backing for AI is further underscored by the approval of the IndiaAI mission, the ₹10,372-crore initiative creating computing infrastructure, multi-modal AI Centers for building LLMs and deploying 10,000 GPUs. This massive program seeks to promote innovation, nurture startups, and contribute to establishing India as a leading player in the AI landscape.

OpenAI's Sam Altman on India's Potential

OpenAI chief Sam Altman, during the meeting with Union IT Minister Ashwini Vaishnaw, mentioned that "an extraordinary market for AI" awaits India. Altman highlighted that India is the second biggest market for the AI behemoth, and the country must take a leadership role in spearheading driving innovation related to AI.

Altman emphasised India's expanding role in AI innovation, noting that developers are actively creating tools built on ChatGPT. He noted that Indian users of OpenAI's platforms have tripled in the past year, reflecting the nation's growing engagement with AI technologies.

"India should be a leader in building small models, especially reasoning models," Altman said, emphasising that while AI training costs will continue to rise exponentially, the returns in intelligence and revenue will also grow significantly. According to him, near-term AI models are already reaching the threshold of being good enough to address critical issues like healthcare and education, sectors where India has much to gain from AI-driven innovation.

India is now OpenAI’s second-largest market, highlighting the country’s fast-growing adoption of AI-powered tools. Altman encouraged India to "do everything within the AI stack," indicating that beyond just using AI, the country should actively build and contribute across the AI value chain.

"It’s amazing to see what India has done so far," he added, acknowledging the strides Indian startups, researchers, and developers have made in AI innovation.

Altman also addressed the rising costs of AI model training, saying that while it’s still expensive, the cost per unit of AI intelligence is falling by a factor of 10 every year. However, he pushed back against the notion that this will reduce the need for AI hardware, implying that the demand for AI infrastructure will continue to grow.

Additionally, he clarified that his past comments on building foundational models were taken out of context, perhaps referring to previous debates on whether India should focus on developing its own large-scale AI models or leverage existing ones.

PM @narendramodi Ji guides us to democratise technology. Sam appreciated PM's vision.@sama pic.twitter.com/zEvBIoo7Z5

— Ashwini Vaishnaw (@AshwiniVaishnaw) February 5, 2025

“In reference to the comment I made in India a few years ago about the cost of building foundational AI models, it was taken out of context. That was a certain time of scaling AI, and I still think that pre-trained foundational AI models are expensive. But, one of the most exciting things that have happened in the industry is that there’s a lot that we’ve done now in the distillation of AI models,” Altman said. “There’s a lot that we’ve done with small models, and reasoning models today are not cheap, but still doable. This can lead to an explosion of creativity, and India should be a leader there.”

Challenges and Global Alliances

Delhi was the final stop of Altman's Asia-Pacific tour, in which he had sessions with lawmakers, AI researchers, and business leaders. His visit comes at a time when India is stepping up its AI ambitions, with initiatives such as Bhashini for AI-based language translation, a Government-backed AI compute infrastructure, and a growing focus on AI regulation, among others.

However, building indigenous AI chips and models isn't without challenges. The Semiconductor industry is ruled by not more than a handful of countries, with the US accounting for more than 70 per cent of the global semiconductor revenue. Setting up a self-reliant semiconductor ecosystem in India will require huge investments in R&D. Building competitive LLMs must therefore ensure access to very big datasets and massive computational power.

To counter these hurdles standing before India, the country is forming alliances with global tech companies. For instance, NXP Semiconductors announced plans to invest around $1 billion in India to double its research and development efforts. Such alliances are expected to boost India's progress in AI chip development and sharpen its abilities in AI research.

Further, India's strong domestic IT sector, worth $250 billion-strong in the economy, and a pool of nearly 5 million programmers serve as a sound base for AI innovation. The AI services market is expected to be valued at $17 billion by 2027, suggesting high growth prospects. Also, massive adoption rates of AI technologies among India's knowledge workers, with 92 per cent reportedly using generative AI at work, underscore the country's readiness to latch on to AI.

Balancing Innovation with Ethical AI Practices

Talking to ETV Bharat, Karnnika A Seth, Advocate & Cyberlaw Expert and Founder of Seth Associates law firm, highlighted the challenges of AI model integrity and security.

“Having an AI in an indigenous model comes with many potential risks. One of them is data privacy leakage, where personal sensitive data could be stolen. There is also the risk of data poisoning, where errors are introduced into datasets, corrupting the final outputs of AI models," she said. "Additionally, AI systems could be vulnerable to backdoor attacks or malicious code insertions. Another concern is reverse engineering, where the structure and parameters of an AI model are extracted, leading to the theft of its outputs."

Seth highlighted the need to focus on ways to mitigate the associated risks with the system. "There are various ways to keep these systems secure and to ensure fairness in their operations. One can implement robust cybersecurity measures such as data encryption, access controls, regular security audits, and continuous monitoring of AI systems," she said, adding that data ethics must also be considered when creating AI models.

“Reducing dependence on foreign processes like Nvidia is crucial, especially when we lack advanced semiconductor technology expertise and are still building our ecosystem. Beyond that, infrastructure is key—we need state-of-the-art, robust facilities to develop these technologies," she added.

Seth also described economic feasibility as another important factor to consider when competing with global brands, which require "a strong supply chain" and the need to take into consideration the complications with semiconductor manufacturing, which include sourcing raw materials and investing in R&D.

“Policy and regulation are also essential," she added. "Without a strong, favourable policy environment to regulate AI and mitigate its risks, we will face significant challenges. Continuous effort is required, and a Public-Private Partnership (PPP) model can help achieve these goals."

Seth also emphasised the importance of focusing on effective data governance to ensure data security, resilience, and the development of an AI-driven ecosystem in India. She noted that without the right framework, data quality and integrity cannot be ensured. She also highlighted the crucial need to build AI ethically, adhering to principles such as transparency, fairness, justice, inclusiveness, and sustainable development.

“To ensure ethical AI is created and deployed, continuous monitoring, audits, and compliance with data protection laws are necessary. Having the right leadership, skilled manpower, and collaboration is key," she said. "Once again, the PPP model is crucial—industry, experts, and the government must work together to leverage AI for the country’s growth and a 'Viksit Bharat'."