As the United States and China battle to dominate artificial intelligence, this week India attempted to highlight other pathways to navigate the silicon surge.
Billed as the first high-level AI gathering to be held in the Global South, the India AI Impact Summit has given the world's most populous country a stage to promote itself as a global AI player, broadening the AI conversation to include countries in Latin America, Africa and beyond.
“Long term, it’s good for the world that AI is not just viewed as a race between the U.S. and China, and I think that India is right now the player that most confidently says, ‘We reject this dynamic,’” said Jakob Mökander, director of science and technology policy at the Tony Blair Institute for Global Change.
As the event's “impact” branding suggests, the summit highlighted how countries can adopt and adapt increasingly powerful AI systems to their own needs and industries.
“Every country will want to chart their own AI destiny,” Michael Kratsios, director of the White House Office of Science and Technology Policy and the leader of the U.S. delegation at the summit, told NBC News. “They each have unique characteristics about their culture, their language, their traditions, the way that they want to use AI.”
As part of the event, Kratsios announced a series of initiatives to increase America's global engagement on AI, including an AI-focused Peace Corps program and new World Bank funding for countries to buy AI systems.
The five-day summit in New Delhi, hosted by Prime Minister Narendra Modi, opened with its share of hiccups. With over 250,000 registered attendees, the summit was plagued early this week with complaints of overcrowding, long lines, visa issues and traffic disruptions, and there were some major no-shows, including a last-minute cancellation from Nvidia CEO Jensen Huang.
Microsoft founder Bill Gates, who has faced questions over his ties to convicted sex offender Jeffrey Epstein, also pulled out hours before his keynote address on Thursday, saying in a statement that he wanted “to ensure the focus remains on the AI Summit’s key priorities.”
Later in the day, U.S. rivals Sam Altman of OpenAI and Dario Amodei of Anthropic held hands with the people on either side of them but not with each other in what was supposed to be a show of unity with Indian Prime Minister Narendra Modi and other tech leaders.
Earlier in the week, an Indian university was reportedly asked to leave the summit after a staff member passed off a robotic dog developed by Chinese company Unitree as one the university had developed.
“The Modi government has made a laughing stock of India globally, with regard to AI,” the opposition Congress party said in a post on X, noting that India’s information technology minister had shared the erroneous report and then deleted it.
But the summit still brought some wins for India, two of whose biggest conglomerates, Reliance and Adani, pledged a combined $210 billion in investment in domestic AI and data infrastructure (compared with the more than $630 billion that U.S. tech giants are expected to spend this year). OpenAI signed a partnership deal with the Mumbai-based Tata Group, while Anthropic announced one with Infosys and opened an office in its home city of Bangalore.
“The solutions presented here — in agriculture, security, assistance for persons with disabilities, and addressing the needs of multilingual populations — are powerful examples of Made in India strength and India’s innovative capabilities,” Modi said in a speech Thursday.
Attendees said that while this year’s summit was accessible to far more people than previous ones, they noted there were fewer of the top government and business leaders who actually make policy. The summit was attended by at least 20 heads of state and government, including French President Emmanuel Macron and Brazilian President Luiz Inácio Lula da Silva, along with tech executives including Google Chief Executive Sundar Pichai and Microsoft President Brad Smith.
The world’s two biggest AI powers, the U.S. and China, did not send heads of state. For China, which has a contentious but warming relationship with India, the summit coincided with the Lunar New Year, the country’s biggest holiday.
As the highest-ranking U.S. official at the event, the White House's Kratsios emphasized the need for countries to eschew strict technocratic oversight of AI, echoing remarks from Vice President J.D. Vance at last year's gathering in Paris.
“AI governance must focus on the particular needs and interests of particular people, and so it must be local,” Kratsios said in a speech Friday. “AI adoption cannot lead to a brighter future if it is subject to bureaucracies and centralized control.”
As part of his speech Friday, Kratsios announced the National Champions Initiative, which aims to help AI-related companies from partner nations create closer links with American AI ventures.
“When we say that we're trying to export the American AI stack, it doesn't mean that it's 100% American content and nothing else is in there,” Kratsios told NBC News. “There are so many countries around the world that have great national AI champions themselves and that excel at certain layers in the AI stack.”
The Trump administration is highly interested in outcompeting China, which has long courted international partners on AI. On Friday, India formally joined Pax Silica, a U.S.-led international coalition launched in December aimed at building a resilient supply chain for critical minerals.
The U.S.-China competition has countries such as India wary of being caught in the middle. Sriram Krishnan, senior White House policy adviser on AI, drew some backlash with his comments this week that the American AI stack should be the “bedrock” its allies build on.
Critics said India should build its own foundational AI models to avoid being too dependent on the U.S. Though India is far behind the U.S. and China, which together control about 85% of global AI computing power, its digital public infrastructure — such as internet connectivity, digital payments and digital ID — is “better than most of the developed world,” Mökander, from the Tony Blair Institute, said.
“They are quite proud about it, that it’s sort of a third way between China’s open-source and closed-source U.S. AI,” he said.
The global AI summit series has evolved significantly since the first one was held in Bletchley Park, England, in 2023, attended by a small group of government and business leaders. While the first summit was focused mainly on the existential risks of frontier AI, this year’s event “takes a very expansive view of what safety means,” said Amlan Mohanty, a fellow at Carnegie India and adviser to the government on AI policy issues.
“It’s no longer only about malicious use, or cybersecurity risks or national security risks,” he said. “It’s about, how can we ensure that we have sufficient data around economic transformation, jobs, impact on labor transitions to be able to make useful policy changes?”
Mohanty said that thinking was reflected in two voluntary commitments made at the summit: one on using data to assess the economic impact of AI, and another on improving the performance of AI models across different languages and cultural contexts.
“It is critical to understand what works, what doesn’t, and who benefits so AI applications can be designed to maximize social benefits and mitigate unintended harms," said Iqbal Dhaliwal, global executive director of the Abdul Latif Jameel Poverty Action Lab at MIT.
“The majority of the world’s population lives in the Global South,” Dhaliwal said in emailed comments, noting that India represents a sixth of the global population. “Hosting the Summit here has allowed the conversation to center the issues and use cases that will affect these billions of people.”
