Introduction

In May 2025, the global venture capital firm BOND released a report titled Trends – Artificial Intelligence. It uses a spectrum of data points to argue that no other technology in history has scaled as fast as artificial intelligence (AI) and outlines how AI is reshaping economic structures, business models, and user behaviour. AI has driven steep upward trends across various indicators—ranging from its projected contribution of US$15.7 trillion to global Gross Domestic Product (GDP) by 2030,^[1] to the tens of billions of devices it has powered since 2020 and its expanding user base. Between October 2022 and April 2025, ChatGPT users alone went from zero to 800 million, and the platform crossed US$4 billion in revenue,^[2] making it the tech sector’s biggest ever overnight-success story.^[3]

This white paper identifies eight key takeaways from the BOND report:

1. Strengthening AI sovereignty can enhance geopolitical influence.
2. As the power and adoption of AI tools increase, underlying infrastructure and talent must be fortified.
3. The escalating demand for compute creates new opportunities for innovation.
4. The production, supply, and availability of high-performance AI chips will be central to states’ AI leadership.
5. Data centres are proliferating, but capacity remains uneven between developing and advanced economies.
6. AI economics show rising revenues but also soaring costs, which limit profitability.
7. As AI disrupts industries, present and future talent must be equipped for a new world of work.
8. Satellite internet may drive the next wave of AI participation.

These eight phenomena impact India just as India, in turn, influences them. The BOND report points out, for example, that India has the largest share of monthly active users (MAUs) of the ChatGPT mobile app and the third largest for DeepSeek.^[4] It also ranks Reliance Industries among the world’s top 25 telecom and tech companies by market capitalisation.^[5]

This paper describes how each of the eight trends listed above is unfolding globally, and explores their implications for the Indian government, regulators, businesses, and markets. It concludes that India could become an AI shaper—defining norms, setting agendas, innovating responsibly, and producing world-class tech talent and solutions.

Takeaways from the BOND Report

1. Strengthening AI sovereignty is a pathway to greater geopolitical influence.

• AI is emerging as a tool for building national resilience and geopolitical influence,^[6] with the global AI race increasingly being defined by US–China rivalry. The US leads in model innovation and custom silicon, while China advances through open-source development, state coordination, and scaled infrastructure.^[7] In 2025, China released three major open-source models: DeepSeek-R1, Qwen-32B, and ERNIE 4.5.^[8] It also has the world’s largest industrial robot base,^[9] generates vast data troves, and is pushing for indigenous innovation.^[10] Local AI platforms dominate China’s domestic AI use, whereas US models lead globally, highlighting a growing platform and regulatory divide.^[11] Meanwhile, the US developer ecosystem is expanding rapidly—NVIDIA’s base grew sixfold to 6 million over seven years, more than doubling in just the last four.^[12] Similarly, Google’s AI ecosystem grew fivefold year-on-year to 7 million developers, underscoring the scale and momentum of US-led AI platforms.^[13]

• States have become more attuned to the imperative of producing AI using their country’s own infrastructure, data, models, and business networks.^[14] ‘Sovereign AI’ encompasses both physical and data infrastructure—ranging from local cloud and compute infrastructures to foundation models, such as large language models (LLMs) developed by local teams and trained on local datasets to ensure inclusion.

• India must strengthen its AI Mission to build sovereign capabilities and reduce external dependence. This includes developing a strong AI ecosystem that fosters innovation and enables effective regulation. For both government and businesses, there is a clear opportunity to build computing capacity. While the government is making 18,000 GPU-based AI servers available to enterprises at subsidised rates,^[15] efforts to build indigenous GPUs could be accelerated.

• Creating foundational LLMs is another core element of AI sovereignty. Sarvam AI^[16] has been selected to build India’s first LLM, but alternative models must also be encouraged. Private players should focus on developing more localised AI solutions and applications. For instance, AI4Bharat is developing open-source language models for Indian languages,^[17] while Soket AI,^[18]ai,^[19] and Gan.ai^[20] are building foundational models tailored for India.^[21] The BharatGPT programme, launched by IIT Bombay and Reliance Jio in 2024, aims to drive innovation across a broad spectrum of local products, services, and sectors.^[22] At the international level, India must remain a leading voice on ethical AI at forums like the Global Partnership on AI and the AI Action Summit, while deepening bilateral cooperation on responsible AI.

2. As the power and adoption of AI tools increase, underlying infrastructures and talent must be fortified. 

• AI adoption is accelerating across consumers, enterprises, and governments, outpacing past waves of technological change. ChatGPT reached 800 million users in 30 months, generating nearly US$4 billion in revenue and 20 million subscribers.^[23] The number of powerful AI models is growing by 167 percent annually,^[24] AI-related job demand is surging, and traditional IT hiring is declining.^[25]

• AI tools are evolving from assistants to service providers.^[26] New AI agents can reason, act, and complete multi-step tasks, managing workflows and executing commands via natural language. A race has now begun to build agentic interfaces, enterprise copilots, autonomous systems, and sovereign models. As the BOND report notes, AI is already transforming sectors such as software, healthcare, law, finance, and customer service, while also probing new frontiers in medical discovery, robotics, autonomous research, supply chains, cybersecurity, energy management, and climate monitoring.^[27]

• India has continued to develop a range of AI use cases across health, education, agriculture, financial services, and smart cities, in line with its 2018 National Strategy for Artificial Intelligence.^[28] However, its focus is now expanding to include foundational model development, compute infrastructure, and multilingual AI—reflecting its ambition to shape global AI innovation, not merely adopt it.

• India now faces opportunities to build sovereign AI models, grow its AI-ready workforce, and strengthen its talent pipeline. It is advancing its capabilities through the development of indigenous foundational models—including LLMs and small language models—such as BHASHINI, BharatGen, Sarvam-1, Chitralekha, and Hanooman’s Everest 1.0.^[29] On the skilling front, the FutureSkills programme under the IndiaAI Mission aims to expand access to AI education at the undergraduate, master’s, and Ph.D. levels, helping build a robust tech talent pool.^[30] This is critical, as AI jobs in India are projected to grow significantly—with estimates pointing to 2.3 million new openings by 2027^[31] and around 2.73 million by 2028.^[32] Expanding compute capacity, scaling data centres, and investing in next-generation AI chip development will also be crucial for supporting Indian innovation.

3. The escalating demand for compute creates new opportunities for innovation.

• The bulk of spending in AI LLM development is dominated by compute—specifically, the resources required to train and run AI models. Training costs remain extraordinarily high and are rising fast, often exceeding US$100 million per model.^[33] As AI becomes more sophisticated and widely used, compute demand is increasing exponentially.^[34] As NVIDIA CEO Jensen Huang has noted, “Reasoning AI agents require orders of magnitude more compute.”^[35] A new ecosystem of infrastructure specialists and service providers is emerging to meet this demand. For instance, CoreWeave is one of the fastest-scaling cloud GPU providers,^[36] while Oracle has positioned itself as a GPU-rich cloud platform offering AI-specific services.^[37]

• There is much room for new public-private partnerships (PPPs) as India seeks diversified commitments for 50,000 GPUs.^[38] Bilateral agreements can further bolster India’s AI capabilities. The US-India TRUST initiative offers a model for advancing AI infrastructure through joint government and private sector innovation. Similarly, the India-France Declaration on Artificial Intelligence aims to catalyse the “emergence of digital public infrastructures for AI by promoting shared objectives and the development of common resources in the field of data, personal data protection, open-source tools and capacity building.”^[39] Moreover, as global tech giants expand their AI and cloud infrastructure in India—Microsoft, for instance, has announced a US$3-billion investment^[40]—Indian conglomerates could also explore opportunities to scale related investments and innovation.

4. The production, supply, and availability of high-performance AI chips will be central to states’ AI leadership.

• Semiconductors are fundamental to AI development and deployment, with capital expenditure increasingly directed toward specialised chips such as GPUs, TPUs, and AI accelerators.^[41] The high demand for NVIDIA’s AI chips, for example, drove its revenue up 78 percent year-on-year, reaching US$39 billion in 2025.^[42] Semiconductors and critical minerals are no longer just economic assets—they are strategic tools of national resilience and geopolitical influence.^[43] The US is strengthening partnerships with trusted allies such as Japan, South Korea, and the Netherlands to reduce reliance on Chinese supply chains.^[44] However, China remains the dominant global supplier of rare-earth elements.^[a] Meanwhile, Taiwan produces over 60 percent of all semiconductors and nearly 90 percent of the most advanced chips, cementing its central role in the global supply chain.

• India is working to advance its semiconductor capacity, with five plants under construction at the time of writing. These facilities will support AI chip production and strengthen India’s role in global electronics and AI supply chains.^[45] However, greater effort is needed to advance the Semiconductor Mission. This includes expanding critical mineral supply chains, nurturing specialisation across key nodes of the value chain, and offering incentives to attract both global and domestic players. India must also focus on setting up robust supply chains to support local electronics and semiconductor manufacturing. Accelerating approvals and funding for additional units under the mission will be key to scaling operations and competing globally.

5. Data centres are mushrooming, but there is an asymmetry in data centre capacity between emerging or developing nations and advanced economies. 

• In 2024, global capital expenditure on data centres reached US$455 billion,^[46] driven primarily by the growing compute demands of AI. Data centres form the backbone of the AI ecosystem, enabling the training and deployment of increasingly complex models. New data centres are being built at unprecedented speed and scale,^[47] but a wide capacity gap persists between developing economies and countries like the US, European Union, and China. While developing economies (excluding China) account for over 50 percent of global internet users, they host less than 10 percent of the world’s data centre capacity.^[48] Across all regions, however, the massive energy consumption of data centres remains a challenge.^[49]

• India has the potential to become a hub for data centre development, but this will require coordinated action across stakeholders. Regulators must ensure clear and consistent policies on data transfers, privacy, and localisation. In April 2025, the European Data Protection Supervisor (EDPS) disclosed that it had denied a February 2024 request from the European Investment Bank (EIB) to transfer contact data to India, citing concerns over the adequacy of India’s data protection framework.^[50] With the Digital Personal Data Protection Act, 2023 still pending enforcement and its accompanying rules yet to be finalised, uncertainty persists around the robustness of India’s data regime. This could undermine user confidence and deter international data flows.

• Access to capital and incentives should support the high upfront costs of data centres. Central and state governments must streamline processes for acquiring land, power, and connectivity to facilitate their development. The ability to attract foreign investment will depend on favourable tax regimes, ease of doing business, and strong PPPs. Finally, India offers a compelling destination for clean energy-powered data centres, given its abundant renewable energy resources and evolving green energy infrastructure.^[51] Going forward, integrating renewable energy solutions will be essential for managing the growing energy demands of data centres.

6. The economics of AI are marked by rising revenues but also spiralling costs which constrain profitability.

• The economics of AI are being shaped by massive capital investments, R&D spending, and fierce competition to meet surging global demand. Tech companies are investing billions in capex to build infrastructure that can capture, process, and monetise data in real time.^[52] While AI adoption and revenues are rising, and inference is becoming cheaper, profitability remains constrained by the high costs of training large models and the rise of open-source alternatives.^[53] Overall, pricing power is eroding.

• This presents a strategic opportunity for India to position itself as both a major AI market and a key player in the global data value chain. With one of the world’s largest internet and consumer application user bases, India holds immense potential for AI deployment and localised data generation. By leveraging its strong developer ecosystem and expanding digital infrastructure, India could lead the deployment of open-source models. Realising this potential will require investments in sovereign AI capabilities and novel public-private collaborations. Initiatives like Sarvam AI and AI4Bharat’s IndicBERT and Chitralekha exemplify homegrown efforts to build multilingual, open-source models tailored to Indian contexts.

7. As AI disrupts industries, present and future talent pools must be equipped for a new world of work. 

• AI is changing the world of work. Along with the growth of physical automation, cognitive automation is now on the rise,^[54] with AI systems increasingly capable of reasoning, creating, and problem-solving. This has widespread ramifications. Industries involving large numbers of knowledge workers, content creators, and manufacturing processes are particularly vulnerable to disruption. In the US, the percentage of firms using AI across all industries rose by 5 percent between 2023 and 2025, and by nearly 15 percent in professional, scientific, and technical sectors.^[55] Workers using AI are found to be 14 percent more productive than others.^[56] Since 2018, AI-related job postings have increased by 448 percent, while non-AI IT job postings have declined by 9 percent.^[57]

• India can take several steps to build an AI-ready workforce. Under the IndiaAI Mission, the FutureSkills pillar is promoting new academic programmes and supporting the creation of data and AI labs in Tier 2 and 3 cities.^[58] Centres of Excellence for AI are also being established at top engineering institutes like the IITs, while dedicated AI-focused universities—such as the Universal AI University in Maharashtra^[59]—represent a nascent trend. Given the rising tide of AI jobs, higher education institutes (HEIs) need to partner more closely with industry to align AI curricula and projects with market requirements. Training AI educators and investing in advanced infrastructure, such as AI labs, will be essential. Larger numbers of open-access AI skilling courses for different levels of learners should be made available through online learning platforms. Finally, STEM education must be revitalised, and an interest in AI inculcated during learners’ school-going years.

8. Satellite internet could drive the next wave of AI participation 

• Global internet penetration has risen to 68 percent—up from 16 percent nearly two decades ago^[60]—with 83 percent of urban dwellers and 48 percent of rural residents online.^[61] Yet, 2.6 billion people remain offline altogether.^[62] Low-cost satellite internet is helping close this gap, expanding global connectivity and unlocking new opportunities in the AI era.^[63] Satellite networks like SpaceX’s Starlink, which now has over 5 million subscribers, are rapidly extending coverage to underserved regions.^[64] As connectivity improves, AI adoption is expected to accelerate further. SpaceX’s leadership in the global orbital launch market is driving this momentum, positioning satellite internet as a key driver of the next wave of global AI participation.^[65]

• India should harness satellite communication (satcom) to boost internet services and bridge digital divides, particularly in rural and remote areas. With Starlink recently receiving its licence to launch satcom services, this process is set to begin. Accordingly, India must evolve policies for satellite internet integration and begin building a domestic satellite and ground equipment industry that upholds national security and data privacy standards. Recent regulatory updates addressing inter-satellite links, mobility, and data security reflect this shift in priorities. Additionally, Indian businesses should pursue partnerships with foreign firms to co-develop solutions and build capacity in this strategically important sector. Domestic players such as Bharti Airtel^[66] and Reliance Jio^[67] have already begun to enter into such collaborations.

Conclusion

AI is reshaping the global technological and economic order, and India has the potential to be a decisive force in this transformation. As this white paper outlines, building sovereign AI capabilities, expanding compute infrastructure, and nurturing talent pipelines are critical. India’s large consumer base, vibrant developer ecosystem, and rising digital footprint offer both a testbed and a launchpad for innovative AI solutions. Realising this potential will require coordinated action across an array of stakeholders. With foresight and focus, India can evolve from an AI adopter to an AI shaper—defining norms, setting agendas, innovating responsibly, and leading inclusively.

Endnotes

^[a] Rare-earth elements (REEs) are crucial components in the manufacturing of computer chips and other advanced technologies, including AI chips.

^[1] “Sizing the Prize: PwC’s Global Artificial Intelligence Study – Sizing the Prize,” PwC, 2017, https://www.pwc.com/gx/en/issues/artificial-intelligence/publications/artificial-intelligence-study.html

^[2] “Trends – Artificial Intelligence,” Slide 19.

^[3] “Trends – Artificial Intelligence,” Slide 7.

^[4] “Trends – Artificial Intelligence,” Slide 316, 317.

^[5] “Trends – Artificial Intelligence,” Slide 277.

^[6] “Trends – Artificial Intelligence,” Slide 272.

^[7] “Trends – Artificial Intelligence,” Slide 337.

^[8] “Trends – Artificial Intelligence,” Slide 261.

^[9] “Trends – Artificial Intelligence,” Slide 288.

^[10] “Trends – Artificial Intelligence,” Slide 271.

^[11] “Trends – Artificial Intelligence,” Slide 2.

^[12] “Trends – Artificial Intelligence,” Slide 38.

^[13] “Trends – Artificial Intelligence,” Slide 39.

^[14] “Trends – Artificial Intelligence,” Slide 77.

^[15] “India’s Own GPU Could Arrive in 3-5 Years, 18,000 AI Servers to Go Live Soon, Says Ashwini Vaishnaw,” The Times of India, February 6, 2025, https://timesofindia.indiatimes.com/technology/tech-news/indias-own-gpu-could-arrive-in-3-5-years-18000-ai-servers-to-go-live-soon-says-ashwini-vaishnaw/articleshow/117958923.cms .

^[16] “About Us,” Sarvam AI, https://www.sarvam.ai/about-us.

^[17] “Building AI for India!,” AI4Bharat, https://ai4bharat.iitm.ac.in/.

^[18] “Building AGI with a Conscience,” Soket, https://soket.ai/.

^[19] “Overview,” Gnani AI, https://www.gnani.ai/about-us/overview/.

^[20] “After Sarvam.ai, Soket AI, Gnani.ai, Gan.ai to Now Build Foundation AI Models, Says IndiaAI Mission,” The Economic Times, May 30, 2025, https://economictimes.indiatimes.com/tech/artificial-intelligence/after-sarvam-ai-soket-ai-gnani-ai-gan-ai-to-now-build-foundation-ai-models-says-indiaai-mission/articleshow/121518886.cms?from=mdr.

^[21] “Airtel, Jio Leverage Artificial Intelligence to Offer Better, Cost-Efficient Customer Experience,” Business Today, October 23, 2018, https://www.businesstoday.in/industry/telecom/story/airtel-jio-ai-artificial-intelligence-customer-experience-150785-2018-10-23.

^[22] “Reliance Jio and IIT Bombay to Launch BharatGPT Program,” IndiaAI, January 5, 2024, https://indiaai.gov.in/news/reliance-jio-and-iit-bombay-to-launch-bharat-gpt-program.

^[23] “Trends – Artificial Intelligence,” Slide 19.

^[24] “Trends – Artificial Intelligence,” Slide 18.

^[25] “Trends – Artificial Intelligence,” Slides 332-334.

^[26] “Trends – Artificial Intelligence,” Slide 89.

^[27] “Trends – Artificial Intelligence,” Slides 231-245.

^[28] NITI Aayog, National Strategy for Artificial Intelligence, June 2018, https://www.niti.gov.in/sites/default/files/202303/National-Strategy-for-Artificial-Intelligence.pdf.

^[29] Ministry of Electronics and IT, “India’s AI Revolution: A Roadmap to Viksit Bharat,” March 6, 2025, https://www.pib.gov.in/PressReleasePage.aspx?PRID=2108810.

^[30] “IndiaAI FutureSkills,” India AI Mission, https://indiaai.gov.in/hub/indiaai-futureskills.

^[31] “2.3 Mn Job Openings by 2027 in India's AI Sector, Offers Opportunity to Reskill over 1 Mn Workers: Bain & Company,” The Economic Times, March 10, 2025, https://economictimes.indiatimes.com/jobs/fresher/2-3-mn-job-openings-by-2027-in-indias-ai-sector-offers-opportunity-to-reskill-over-1-mn-workers-bain-company/articleshow/118850403.cms?from=mdr.

^[32] “AI Could Create 2.73 Million Jobs in India by 2028: ServiceNow Study,” Business Standard, November 13, 2025, https://www.business-standard.com/industry/news/ai-could-create-2-73-mn-jobs-in-india-by-2028-servicenow-study-124111301863_1.html.

^[33] “Trends – Artificial Intelligence,” Slide 100.

^[34] “Trends – Artificial Intelligence,” Slide 134.

^[35] “Trends – Artificial Intelligence,” Slide 116.

^[36] “Trends – Artificial Intelligence,” Slide 158.

^[37] “Trends – Artificial Intelligence,” Slide 158.

^[38] Peerzada Abrar, “India Doubles Down on AI with Rs 10,000 Crore Fund, 50,000 GPU Ambition,” Business Standard, June 5, 2025, https://www.business-standard.com/industry/news/india-doubles-down-on-ai-with-rs-10-000-crore-fund-50-000-gpu-ambition-125060500096_1.html.

^[39] India-France Declaration on Artificial Intelligence, Ministry of External Affairs, Government of India, February 12, 2025, https://www.mea.gov.in/bilateral-documents.htm?dtl/39034/IndiaFrance_Declaration_on_Artificial_Intelligence_February_12_2025.

^[40] Veena Mani and Shilpa Phadnis, “Microsoft to Invest $3 Billion in India, Boost AI, Cloud Infrastructure,” The Times of India, January 8, 2025, https://timesofindia.indiatimes.com/business/india-business/microsoft-to-invest-3-billion-in-india-boost-ai-cloud-infra/articleshow/117035643.cms.

^[41] “Trends – Artificial Intelligence,” Slides 95, 157-168.

^[42] “Trends – Artificial Intelligence,” Slide 160.

^[43] “Trends – Artificial Intelligence,” Slide 272.

^[44] “Trends – Artificial Intelligence,” Slide 272.

^[45] Ministry of Electronics and IT, Government of India, “India’s AI Revolution: A Roadmap to Viksit Bharat”.

^[46] “Trends – Artificial Intelligence,” Slide 118.

^[47] “Trends – Artificial Intelligence,” Slide 120.

^[48] “Trends – Artificial Intelligence,” Slide 125.

^[49] “Trends – Artificial Intelligence,” Slide 125.

^[50] Tanusha Tyagi, “The Adequacy Dilemma: India's DPDPA and the GDPR,” Observer Research Foundation, June 18, 2025, https://www.orfonline.org/expert-speak/the-adequacy-dilemma-india-s-dpdpa-and-the-gdpr.

^[51] “World Is Looking Up to India to Set Up Data Centres: Piyush Goyal,” The Economic Times, January 14, 2025, https://economictimes.indiatimes.com/tech/technology/world-is-looking-up-to-india-to-set-up-data-centres-piyush-goyal/articleshow/117243432.cms?from=mdr.

^[52] “Trends – Artificial Intelligence,” Slides 94-116.

^[53] “Trends – Artificial Intelligence,” Slides 248-298.

^[54] “Trends – Artificial Intelligence,” Slide 324.

^[55] “Trends – Artificial Intelligence,” Slides 328-329.

^[56] “Trends – Artificial Intelligence,” Slide 331.

^[57] “Trends – Artificial Intelligence,” Slide 332.

^[58] “IndiaAI FutureSkills”.

^[59] “Qatar’s Arab Center for Artificial Intelligence Partners with Universal AI University for Advancement of AI,” The Tribune, June 3, 2025, https://www.tribuneindia.com/news/business/qatars-arab-center-for-artificial-intelligence-partners-with-universal-ai-university-for-advancement-of-ai/.

^[60] “Trends – Artificial Intelligence,” Slide 311.

^[61] “Trends – Artificial Intelligence,” Slide 313.

^[62] “Trends – Artificial Intelligence,” Slide 309.

^[63] “Trends – Artificial Intelligence,” Slide 309.

^[64] “Trends – Artificial Intelligence,” Slide 320.

^[65] “Trends – Artificial Intelligence,” Slide 318-322.

^[66] “Airtel Announces Agreement with SpaceX to Bring Starlink’s High-Speed Internet to its Customers in India,” Airtel, https://www.airtel.in/press-release/03-2025/airtel-announces-agreement-with-spacex-to-bring-starlinks-high-speed-internet-to-its-customers-in-india/.

^[67] “Jio Partners with Elon Musk's SpaceX to Bring Starlink’s High-Speed Satellite Internet to India,” Business Today, March 12, 2025, https://www.businesstoday.in/technology/news/story/jio-partners-with-spacex-to-bring-elon-musks-starlinks-high-speed-satellite-internet-to-india-467621-2025-03-12.

• Artificial Intelligence
• India
• ai action summit
• AI adoption
• ai bilateral cooperation
• ai ecosystem in india
• ai in india
• ai llm development
• ai platforms
• AI sovereignty
• BharatGPT programme
• global AI race
• global partnership on ai
• llm platforms
• llm platforms india
• semiconductor mission
• US China AI rivalry
• us china ai war
• us india trust initiative

The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.