Updated Aug 14

AI Strategy Debate: Build New or Use Existing?

India's AI Strategy Showdown: Building New Models vs. Leveraging Existing Ones

A heated debate is brewing over India's approach to AI development. Perplexity AI CEO Aravind Srinivas and Infosys co‑founder Nandan Nilekani clash on the best path forward. While Nilekani suggests focusing on applications using existing LLMs, Srinivas envisions an ISRO‑like feat by building new, cost‑effective AI models. With Google Research India's Manish Gupta siding with Srinivas, the debate highlights vital implications for India's technological, economic, and cultural future.

Introduction to India's AI Strategy Debate

India's Artificial Intelligence (AI) strategy is currently the subject of intense debate among the country's leading technology thinkers. This discourse is centered around a pivotal choice: whether India should focus on enhancing AI applications using existing large language models (LLMs) or pursue the development of its own foundational AI models from scratch. Fueling this debate are divergent views from tech giants such as Nandan Nilekani, Infosys co‑founder, and Aravind Srinivas, CEO of Perplexity AI.

Nilekani, with his practical stance, advocates for utilizing established AI models to develop useful applications, arguing that creating new foundational models from the ground up is excessively capital‑intensive for India. He draws parallels to the Aadhaar project, suggesting that leveraging existing technology and infrastructure for broad applications can yield significant benefits efficiently.

Conversely, Srinivas envisions a more ambitious pathway, encouraging India to embark on building indigenous foundational AI models. He likens this approach to an 'ISRO‑like feat' in AI, underscoring the potential for India to achieve self‑sufficiency and innovation in AI technology. His perspective is supported by Manish Gupta from Google Research India, who emphasizes the importance of creating AI capabilities at the foundational level to ensure technological independence and competitiveness.

This debate is not just theoretical but practical, with implications for India's position in the global AI landscape. Should India succeed in building AI models from scratch - particularly those catered to Indic languages - it could boost its technological sovereignty while fostering innovation tailored to its unique linguistic and cultural context. However, this approach requires significant investment in AI research, development infrastructure, and skilled workforce.

Nandan Nilekani's Argument for Utilizing Existing LLMs

Nandan Nilekani, a prominent figure in India's tech industry and co‑founder of Infosys, has recently put forth an argument favoring the utilization of existing Large Language Models (LLMs) for advancing AI applications in India rather than engaging in a resource‑intensive journey to develop new foundational AI models. His position stems from a pragmatic analysis that acknowledges the current capabilities and global technological landscape of AI.

Nilekani argues that leveraging existing LLMs can steer India's AI initiatives toward practical implementations and applications that are more immediate and less costly. According to him, the focus should be on how these pre‑existing models can be effectively adapted and integrated into Indian contexts to yield significant economic and social benefits. He sees this approach as a way to harness the power of AI without being bogged down by the substantial financial and infrastructural demands of creating new foundational models from scratch.

Drawing parallels with Aadhaar, India's unique identification system that has transformed governmental and economic processes, Nilekani believes that a similar strategy could be applied to India's AI sector. By utilizing already developed technologies in innovative ways, India can leapfrog into advanced AI applications, thereby achieving substantial progress without the initial overheads associated with building AI from the ground up. This approach also minimizes risks related to long development timelines and uncertain outcomes that are typically associated with pioneering new foundational technology.

Critically, Nilekani's viewpoint acknowledges the importance of building indirect sovereignty and expertise through smart utilization rather than direct creation, which may not always be feasible given India's current resource allocation. This perspective has sparked significant discourse among tech leaders and experts, reflecting a broader debate on the balance between immediate practicality and long‑term innovation in India's AI strategy.

Aravind Srinivas's Vision for Indigenous AI Development

Aravind Srinivas, the CEO of Perplexity AI, proposes a visionary approach for India, aspiring for significant achievements in AI akin to ISRO's success in space technology. This vision emerges as a counter to Nandan Nilekani's strategy, emphasizing the development of India's indigenous AI capabilities. Srinivas argues for a transformative shift by creating foundational AI models domestically, suggesting that such endeavors could lead to global competitiveness and technological independence.

In the landscape of technological advancement, Srinivas envisions that India can achieve a landmark feat similar to ISRO's approach, renowned for its cost‑effective and innovative solutions. He believes that by focusing on indigenous AI development, India could craft AI models tailored to its unique cultural and linguistic needs, thereby capitalizing on a distinct niche unaddressed by current global AI technologies.

Aligning with Srinivas, Manish Gupta from Google Research India also endorses the creation of foundational AI models within India. He stresses the importance of building a strong technological foundation rather than primarily focusing on AI application development using existing large language models (LLMs). The consensus among these leaders underscores the potential of Indian innovation in AI if combined with dedicated infrastructure and research investment.

The implications of pursuing a homegrown AI strategy are multidimensional. Economically, it promises the creation of numerous high‑skilled jobs and the emergence of a robust AI research ecosystem that could substantially contribute to India's tech industry. Furthermore, it could secure India's position as a technological powerhouse by advancing its independence from foreign AI entities, while also ensuring AI solutions that incorporate India's diverse linguistic and cultural tapestry.

Manish Gupta's Alignment with Srinivas's Approach

Manish Gupta, the Director of Google Research India, has expressed his alignment with Aravind Srinivas's approach regarding India's AI development strategy. This alignment signifies a convergence of thought among prominent figures in the AI domain towards the need for India to develop indigenous foundational models rather than solely relying on existing large language models (LLMs).

Gupta's support for Srinivas's vision highlights a growing belief that India's progress in AI should mirror the country's achievements in space technology, such as those by the Indian Space Research Organisation (ISRO). By focusing on creating foundational models, India would not only enhance its technological capabilities but also ensure a strategic independence similar to that seen in its space programs.

The backing from such a reputable source within the industry, like Gupta of Google Research India, underscores the value of innovation and developing AI tailored to India's specific needs. It reflects an understanding that investing in foundational models can ultimately drive economic growth and technological sovereignty, reducing reliance on foreign technologies.

By choosing to develop its own AI models, India could play a crucial role in addressing the nuances of Indic languages and diverse cultural contexts, thereby carving out a niche in the global AI landscape. This direction could foster a stronger AI ecosystem within the country, promoting sustainable growth and positioning India as a leader in the AI field.

Furthermore, aligning with Srinivas's approach signifies a strategic vision for India where creating its own path in AI development is seen as essential for long‑term success, ensuring competitive advantage on the global stage and reinforcing the country's standing as a technological powerhouse.

Practical Implications of Competing Strategies

The debate between different AI development strategies in India holds important practical implications for the country's future. On one hand, adopting the approach advocated by Nandan Nilekani, which focuses on leveraging existing large language models (LLMs) to create applications, presents an immediate cost‑effective solution. This strategy aligns with minimizing initial investments and utilizes the already available technological infrastructure. However, the trade‑off might involve becoming overly dependent on foreign technologies and potentially missing out on developing deeper technical expertise within the country itself.

Conversely, Aravind Srinivas's proposition of developing indigenous AI models from the ground up suggests a longer‑term vision with transformative potential. Drawing a parallel to India's accomplished feats in space technology through ISRO, this strategy emphasizes innovation under resource constraints. If successful, it could not only lead to cost‑effective AI solutions that cater more closely to India's unique requirements, such as supporting Indic languages, but also help establish India as a global AI powerhouse. Moreover, this path might enhance the nation's technological sovereignty, reducing reliance on non‑domestic models and fostering significant advancements in AI research and development.

Supporting arguments for indigenous model development consider the infrastructural investments necessary to sustain AI growth. As both strategies necessitate robust computational infrastructure and AI cloud resources, prioritizing these investments would provide a foundational backbone that supports broader AI initiatives. Developing these capacities internally could propel India toward achieving significant economic gains, with predictions of potentially creating a trillion‑dollar opportunity within the tech sector by 2030.

Ultimately, the choice between these contrasting strategies involves weighing immediate benefits against long‑term aspirations. The decision has far‑reaching implications not just economically, but also culturally and socially, with the potential to reshape India's position on the global technological stage. By carefully navigating this strategic crossroads, India could effectively leverage its rich talent pool and innovative spirit to drive substantial progress in AI.

Feasibility of Achieving an 'ISRO‑like' AI Feat

India has the potential to achieve significant milestones in AI development, akin to the achievements of ISRO in space technology. Inspired by ISRO’s model of cost‑effective innovation, India could leverage its technical expertise to build indigenous AI models that cater specifically to its diverse needs, particularly in Indic languages. The success of such an endeavor would not only establish technological sovereignty but also enable India to become a key player in the global AI landscape.

A central debate in this pursuit is whether India should focus on creating its own foundational AI models or maximize existing Large Language Models (LLMs). Tech leaders like Aravind Srinivas advocate for building from scratch to enhance global competitiveness, whereas Nandan Nilekani suggests leveraging current models for immediate application, emphasizing practicality over capital‑intensive developments.

The discussion is further enriched by various experts, public opinions, and geopolitical dynamics. High‑profile figures like Manish Gupta of Google Research India align with Srinivas, stressing the importance of foundational models, which contrasts sharply with Nilekani's more pragmatic approach. The public, too, shows divided support, with many rallying behind Srinivas, inspired by the possibilities of technological self‑reliance, while others recognize the pragmatic benefits of Nilekani’s cost‑effective approach.

The implications of choosing a particular AI strategy are profound. Economically, indigenous AI model development could create enormous economic opportunities and high‑skilled jobs, potentially positioning India as an AI powerhouse, much like its space program. Furthermore, such development could ensure technological sovereignty and reduce dependency on foreign entities, boosting India’s status in tech diplomacy.

What’s more, this move could provide significant socio‑cultural benefits by enhancing AI support for Indic languages, thereby promoting digital inclusivity and preserving cultural nuances. However, there’s a risk of widening the digital divide if efforts focus predominantly on urban, English‑speaking areas and neglect rural and regional language communities.

In summary, the strategic direction that India takes in AI development will have lasting impacts across economic, technological, socio‑cultural, and international domains. Balancing these factors will be crucial in crafting a path that maximizes India’s potential and fosters a sustainable and inclusive AI landscape.

Role of India in AI for Indic Languages

India's role in advancing AI technologies for Indic languages stands at a crucial juncture in the country's broader AI strategy discourse. A nation characterized by extensive linguistic diversity and a rich cultural tapestry, India has the unique potential to lead the development of AI solutions that cater specifically to its myriad of spoken languages. The strategic contention among Indian tech leaders, as highlighted in the debate between Perplexity AI CEO Aravind Srinivas and Infosys co‑founder Nandan Nilekani, revolves around whether to develop new foundational AI models or leverage existing ones to meet local needs.

Srinivas, backed by thought leaders like Google Research India's Manish Gupta, underscores the need for indigenous AI models, drawing parallels to India's cost‑effective and globally recognized achievements in space technology through ISRO. Such an approach could propel India into a leading position in the development of AI models that not only serve linguistic needs but also foster technological sovereignty. It would also potentially catalyze economic growth, technological innovation, and job creation within the tech sector, transforming India into a global AI powerhouse.

In contrast, Nilekani argues for maximizing the utility of existing large language models (LLMs) by focusing on applications rather than creating new models from scratch. His pragmatic approach considers the substantial investment required for foundational AI model development as financially burdensome, advocating instead for immediate strategic gains through the optimization of current technologies. This method could offer a quicker path to deploying AI solutions across Indic languages, albeit at the potential cost of technological independence.

Despite differing perspectives, there's consensus on the critical nature of enhancing AI capabilities for Indic languages. Building AI models tailored to local linguistic contexts ensures that digital advancements are inclusive, meeting the needs of a wider demographic. It could bridge the digital divide significantly, enabling access to AI‑driven applications for non‑English speaking populations, thereby advancing digital inclusivity and preserving cultural heritage.

As the international community, including China and the EU, make strides in AI development strategies, India's decision regarding its AI trajectory becomes vital. Successfully navigating this path could significantly influence India's role in global tech diplomacy, offering collaborative opportunities and reshaping technological dynamics within the Asian context. The debate emphasizes a balance between leveraging and innovating, aspiring for a future where India’s diverse languages and cultures are seamlessly integrated into the digital realm through advanced AI capabilities.

Necessary Infrastructure Investments

The discussion around necessary infrastructure investments for AI development in India arises from the ongoing debate over the nation's AI strategy. This debate has gained significant attention due to varying opinions from influential tech leaders. On one side, there is a push to build new foundational models, akin to the success India has seen with ISRO in the space sector, as highlighted by Perplexity AI's Aravind Srinivas. The premise here is that creating indigenously developed AI models would offer India a competitive edge and technological independence, fostering innovation that is deeply rooted in local needs and challenges.

Building new models necessitates substantial investment in infrastructure, especially compute infrastructure and AI cloud resources. Such investments would not only support the creation of these foundational models but also facilitate a broader AI‑driven economic landscape within India. This approach promises to secure high‑skilled jobs and establish a thriving research ecosystem, paving the way for India to become a global AI powerhouse.

On the other side of the debate, as advocated by Infosys co‑founder Nandan Nilekani, is the strategy to leverage existing large language models (LLMs) to build AI applications. This approach is seen as more cost‑effective initially, reducing the capital required for infrastructure development. However, the potential downside might be that it can lead to a technological dependency on AI capabilities developed by tech giants outside India, thereby limiting India's role as an innovator in the AI sector.

Regardless of the strategy adopted, experts agree on the necessity of upgrading India's infrastructure to better support AI development. This includes enhancing computing power and creating robust data ecosystems to handle the vast amount of data AI systems operate on. Furthermore, focusing on AI cloud capabilities will provide the backbone needed for both indigenous AI developments and the efficient deployment of AI solutions based on existing models.

Additionally, the development of AI models tailored for Indic languages presents a unique opportunity for India to address a significant gap in the current AI capabilities. This initiative would not only cater to a diverse linguistic population but also help promote cultural inclusivity and digital accessibility across the country, further bolstering India's position as a leader in AI for development. All these factors underscore the strategic importance of investing in infrastructure to support AI initiatives, ensuring both technological sovereignty and competitiveness on the global stage.

Global and Regional Influences on India's AI Strategy

India's AI strategy is subject to both global and regional influences, with various stakeholders voicing differing opinions on the best approach to AI development. At the heart of the debate is whether India should focus on building new foundational AI models domestically, as advocated by leaders like Aravind Srinivas, CEO of Perplexity AI, or leverage existing large language models (LLMs) for AI applications, as recommended by Infosys co‑founder Nandan Nilekani.

Global events such as China's announcement of a comprehensive AI regulatory framework in December 2024 and the EU‑India Tech Partnership Summit in November 2024, which emphasized collaborative opportunities in AI development, play a crucial role in shaping India's AI strategy. These events highlight the importance of technological diplomacy and international cooperation in advancing AI capabilities.

Regionally, India's diverse linguistic landscape offers a unique challenge and opportunity to develop AI models tailored specifically for Indic languages. This potential for growth is further supported by developments such as Karnataka and Tamil Nadu's unveiling of their respective AI policies and innovation hubs in late 2024, aiming to foster a multi‑layered approach to AI development within the country. With global giants like Google also stepping into the fray by launching AI models supporting 12 Indian languages, the regional push for AI advancements is clear.

Diverse Expert Opinions on the Strategic Divide

The realm of artificial intelligence in India is witnessing a pivotal debate with industry giants split over the country's strategic path towards AI development. Anchoring this discussion are two contrasting viewpoints. Nandan Nilekani, co‑founder of Infosys, suggests focusing on applying artificial intelligence using existing large language models (LLMs) to leverage time and resources efficiently. He draws parallels to approaches like Aadhaar, emphasizing practicality and immediate benefits.

On the other end of the spectrum, Aravind Srinivas, CEO of Perplexity AI, argues for the development of foundational AI models from the ground up. Srinivas likens this to an 'ISRO‑like feat' in AI, citing the Indian Space Research Organisation’s landmark achievements as a testament to the potential of indigenous innovation. This perspective underscores the importance of investing in new foundations to ensure long‑term technological sovereignty and global competitiveness.

Adding more depth to the discourse, Google Research India director Manish Gupta aligns with Srinivas. He advocates for the development of local foundational models, claiming this approach is vital for establishing a robust AI industry in India. Gupta criticizes Nilekani's strategy, asserting that the real value lies in strong foundational capabilities rather than dependence on existing global applications.

These differing opinions encapsulate a broader strategic divide that is not just technical but also profoundly impacts economic, social, and cultural spheres in India. On an economic level, developing proprietary AI models could unleash substantial financial growth and job creation, fostering a thriving research ecosystem. Culturally, models tailored to India's linguistic diversity could enhance digital inclusivity and preserve cultural heritage through advanced language processing capabilities.

The implications of choosing between these strategic paths are immense. While building indigenous models demands significant investment and nurturing of local talent, it promises technological independence and a chance to lead in AI, much like India's success in space technology. Conversely, utilizing existing models can be cost‑effective but may lead India towards technological dependency, potentially stifling innovation.

Public reaction to this debate reveals a divided yet passionate sentiment. Many tech professionals and social media users are vocal in their support for Srinivas's bold vision, drawing motivation from India's capabilities in other high‑tech fields like space exploration. Conversely, a pragmatic segment agrees with Nilekani, recognizing the immediate practicality and lower costs associated with using established LLMs.

Internationally, the strategy India adopts will impact its standing as a technological leader. Success in foundational model development could enhance diplomatic relations, strengthen collaborations, particularly with the EU, and possibly reshape technology dynamics, especially in competition with regional powers like China.

In conclusion, the debate over India's AI strategy is more than a technical consideration; it is a crossroads with far‑reaching consequences. Whether India pursues foundational models or leans into existing global infrastructures, the decision will profoundly influence its economic growth, cultural preservation, and international technological standing in the coming decades.

Public Reactions and Societal Perspective

The public's reaction to India's AI strategy debate has been varied, reflecting diverse societal perspectives. On one end, a significant portion of the public, especially among young tech enthusiasts and professionals, is inspired by Aravind Srinivas's vision of developing India's own foundational AI models. This group sees the potential for India to achieve a monumental success similar to that of ISRO's in the space sector. They believe that developing homegrown models could position India as a global leader in AI and drive significant economic opportunities.

Conversely, there are pragmatic voices that echo Nandan Nilekani's recommendations. These individuals point to the financial and resource‑intensive nature of building foundational models from scratch, instead favoring the use of existing language learning models (LLMs) to develop applications. They argue that this approach would allow India to capitalize on AI advancements without the hefty initial investments, thus appealing to stakeholders concerned with immediate cost‑effectiveness.

Social media platforms, particularly X (formerly Twitter), have become vibrant arenas for discussion and debate, with users frequently reposting and commenting on viewpoints expressed by Srinivas and Nilekani. This has resulted in a wider public engagement, not just among tech professionals but also with those interested in India's broader economic and technological advances.

Overall, while the debate continues, the public sentiment appears to incline towards a balanced strategy. The idea of nurturing indigenous AI models is widely supported, yet there is a general consensus on the importance of leveraging existing resources. This dual approach could potentially maximize India's competitive edge while ensuring efficient utilization of available tools and technologies.

Future Economic, Technological, and Social Impacts

India's path to AI development is at a crossroad, with significant implications for its economic, technological, and social landscapes. On one hand, tech leaders like Aravind Srinivas envision a future where India becomes a global AI powerhouse by developing its own foundational models, akin to ISRO's success in space technology. This approach could lead to substantial economic growth, potentially generating over a $1 trillion opportunity, and establishing India as a leader in AI innovation by 2030. Building these models would not only foster a robust AI research ecosystem but also create thousands of high‑skilled jobs, significantly contributing to the country's tech sector expansion.

In contrast, tech stalwart Nandan Nilekani advocates for a more pragmatic approach, leveraging existing large language models (LLMs) to develop AI applications. While this strategy is more cost‑effective and practical in the short term, it risks increasing technological dependency and potentially undermining India's global competitiveness in the AI sector. The debate highlights a pivotal decision for India: to invest heavily now in developing indigenous capabilities or to optimize immediate resources, thereby delaying foundational growth.

Technological sovereignty is another critical aspect of this debate. By developing its own AI models, India could reduce its reliance on foreign technologies, similar to its achievements in space innovation. This independence is especially relevant given the global race in AI development and the need for technological autonomy. Moreover, India has a unique opportunity to lead in creating AI solutions tailored to Indic languages, addressing a significant gap and serving its diverse linguistic population effectively.

The social and cultural impacts of these strategic decisions are also significant. Indigenous AI models could enhance digital inclusion by providing improved accessibility in local languages, thus promoting digital literacy and preserving cultural nuances. However, focusing solely on urban and English‑speaking populations might exacerbate the digital divide, excluding significant portions of the population from technological advancements. India's strategy must balance these elements to promote equitable digital growth.

Finally, the international relations dimension of India's AI development strategy cannot be ignored. Successful indigenous AI advancements would not only enhance India's position in global technology diplomacy but also open doors for increased collaboration with other countries and regions, such as the European Union. As AI becomes a battleground for technological supremacy, India could emerge as a key player in reshaping Asian and global technology dynamics, particularly in competition with countries like China.

International Relations and Technological Diplomacy

In the current global landscape, international relations are increasingly being shaped by advances in technology and the subsequent need for technological diplomacy. This is particularly relevant in the context of artificial intelligence (AI), where strategic decisions regarding development paths and collaborations can significantly affect national sovereignty, economic growth, and international standing. Countries around the world, including India, are engaged in debates and strategic planning to determine the most effective ways to harness AI's potential for national benefit.

India, with its burgeoning tech sector, faces a critical choice in its AI strategy. The debate centers around whether to build new foundational AI models from scratch or focus on developing applications using existing large language models (LLMs). Proponents of the former, like Perplexity AI's CEO Aravind Srinivas, argue that creating indigenous models would be akin to an "ISRO‑like feat," giving India technological independence and a competitive edge globally. Meanwhile, others like Infosys Co‑founder Nandan Nilekani suggest leveraging existing models for cost‑effectiveness, drawing parallels with India's approach to the Aadhaar project.

The implications of India's decision are vast. On an economic level, the development of homegrown AI models could create a trillion‑dollar opportunity by 2030, fostering job creation and innovation. Technologically, it could reduce reliance on foreign models and position India as a leader in AI, especially in the realm of languages, where there is a compelling opportunity to develop AI systems that cater to Indic languages. Socially and culturally, such developments could preserve and promote local languages and cultural nuances, while also enhancing digital inclusion across India's diverse population spectrum.

On the international front, India's AI strategy can influence its diplomatic engagements and collaborations. Success in AI could enhance India's role in global tech diplomacy, facilitating partnerships in research and development with entities like the European Union. However, the drive to excel in AI also puts India in competition with technological powerhouses like China, which is actively advancing its own AI strategies. This dynamic reshapes geopolitical alignments and technological ecosystems in the Asian region and beyond.

More on This Story

Aug 13, 2026

Team Takeoff in Electrical Estimating: Splitting the Work Without Splitting the Accuracy

tech industry

Aug 6, 2026

The Hidden Compliance Risk in Manual Accounts Payable Workflows

tech industry

Aug 6, 2026

Related News

Oct 28, 2026

LangSmith adopts reference_outputs for creating dataset examples

LangSmith’s October 19–26 updates name expected example outputs reference_outputs. The old outputs alias remains valid, but sending both is rejected.

LangSmithLangChainAI development

Oct 20, 2026

Gemini Enterprise can render MCP Apps as interactive interfaces inside chat

Google’s new developer guide shows how MCP Apps pair a tool with a live interface rendered in Gemini Enterprise. The pattern can shorten multi-step work, but developers must still bind the widget to approved tools, isolate it in a sandbox and verify every consequential server action.

Gemini EnterpriseMCP AppsModel Context Protocol

Apr 27, 2026

OpenAI's Five Principles for AI Development Prioritize Ethical Innovation

OpenAI has laid out its five-principle framework for developing AI responsibly. This includes democratizing AI access, empowering users, fostering universal prosperity, ensuring resilience, and maintaining adaptability. Builders should take note, as these principles could influence AI's role in shaping future tech and policy landscapes.

OpenAIAGIAI ethics