The India AI Impact Summit in New Delhi last month made something visible that rarely surfaces in mainstream AI discourse: power in AI systems ultimately rests on control over data. While the expected outcomes were achieved, the event revealed something more fundamental about where the future of AI is actually being shaped — and it’s not just by the technology itself, but by the data systems and the partnerships required to create and govern them. This realization sits at the heart of our work and that of our community.
What the Summit revealed
The Summit didn't result in a meaningful shift in power, but it did demonstrate where that power sits. Billions of dollars in investments and partnerships were announced, many involving Big Tech companies expanding services and infrastructure in India. Media coverage described the event as a global trade fair for AI investment. At the same time, India used the Summit to position itself as a central player in the emerging AI landscape, particularly as a hub for AI development and deployment in the Global South.
India's attractiveness to the global tech world isn't by chance. India is a highly digitized country, ranking just behind the U.S. and China, thanks to investment in digital public infrastructure. Its digital payment system enables 97 percent of financial transactions. The country has a digital identity system that covers nearly everyone in India and has the second largest internet-user base in the world. In other words, India is swimming in data. AI systems are built on data, and so the actors that control how data is collected, shared, governed, and used play a central role in determining who benefits from AI.
Meanwhile, civil society organizations and digital rights groups raised concerns about the lack of meaningful influence over outcomes. As Jeni Tennison of Connected by Data and others pointed out, participation at the Summit was largely tokenistic, replicating the current systems in which the great majority of the world's countries and people have no say in how AI is developed, deployed, and used. Discussions referenced inclusion, but affected communities had limited influence over outcomes.
If the Summit reflected existing power structures, the Participatory AI Research & Practice Symposium (PAIRS) offered a glimpse of a different model. The number of registrations and the sessions at PAIRS demonstrated the great quantity of work and growing interest in building participatory AI systems. Through dozens of sessions, people shared examples of work to meaningfully build participation into the development, deployment and governance of AI systems.
What the data for development community is actually asking
“The conversation about AI governance and the conversation about data systems are the same conversation. What we're hearing from our network, and what we saw in New Delhi, is that the communities and countries with most at stake are ready to move from dialogue to action.”
- Jenna Slotin, interim CEO, Global Partnership
The conversations people want to have about AI governance aren't primarily about safety frameworks or investment flows. They're about who decides, who benefits, and how governance catches up with deployment.
We see this directly in our own network. AI and emerging technology came through as the biggest topic in our recent open call for sessions ahead of the Global Data Festival and Kenya Space Expo & Conference in Nairobi this June. The appetite is there, and so is a clear sense of what needs to be done.
A recurring theme across submissions is the difficulty of moving from pilot to scale. These concerns reflect what happens when governance hasn't kept pace with deployment, and when promising experiments stall because the data foundations, institutional capacity, or political will aren't there to carry them forward.
There's another tension surfacing in how these conversations are being framed, particularly around young people. Across submissions, youth appear in two quite different ways: as vulnerable populations whose data needs protecting and as agents who should be empowered to shape and build AI systems. What it looks like to be both at the same time, and who gets to decide which framing applies, is one of the more interesting questions the community is beginning to sit with as it moves from ambition to implementation.
From data systems to power, and back again
This is why conversations about AI governance can't be separated from investments in data systems. Countries that lag behind in digital connectivity are also excluded from the global AI conversation. The race to benefit from AI has already widened the global digital divide, it and will continue to do so without deliberate intervention.
Despite the concentration of power and wealth on display at the Summit, there are also signs of growing demand for change. As Mozilla's President Mark Surman observed, beyond the public announcements, there was real hunger from countries, companies, and communities to come together and build AI that is open-source, sovereign, and culturally tailored.
Translating that hunger into change requires more than declarations. It requires sustained investment in unglamorous foundational work: strengthening national data ecosystems, improving data sharing, and advocating for data governance frameworks that put people first. Funding across the development and data-for-good ecosystem is shrinking, even as demand for responsible AI systems continues to grow. No single organization, government, group, or sector can address these challenges alone. Partnership isn't just a throughline, it's really the mechanism.
Across our network, partners are working to address these foundational barriers. While less visible than headline investments, these systems ultimately determine whether the benefits of AI are shared widely and the harms addressed.
What we're building toward
In 2026 and into 2027, there will be no shortage of moments to advance these issues — globally, regionally, and nationally. The G20 and G7, the Global Dialogue on AI Governance alongside the ITU AI for Good Summit in Geneva, the World Data Forum in November, and the next AI Impact Summit in 2027 will each bring together different combinations of governments, industry, and civil society. We'll be watching closely how these processes influence each other, and where genuine openings emerge for the communities and countries that have most at stake to shape the direction of travel.
The Global Data Festival and Kenya Space Expo & Conference in Nairobi this June is our contribution to that momentum. It will bring together colleagues from governments, civil society, the private sector, and research communities from all around the world. The agenda is being shaped to give focus to some of the questions our network is already asking: how governance catches up with deployment, how pilots become policy, and how the people most affected by AI systems get a genuine say and influence. Over four days, the focus will be on building new partnerships, exchanging practical solutions, and promoting a shared agenda on data and AI for sustainable development.
As attention continues to focus on big players and high-profile announcements, the message from our network remains consistent: AI innovation must be matched by investment in the data systems it depends on. Who controls those systems—including how they are built, governed, and made accountable—is where the real work of AI governance happens. That's what we're taking from New Delhi, and where the conversation picks up in Nairobi.
