Notes from the AI Governance Center: Outcomes of the India AI Impact Summit

In February, the AI world gathered in India for the latest installment of the AI Summit series. Dubbed the AI Impact Summit, this weeklong event proved to be an important showcase for industry, governments and international organizations to provide direction and share evolving best practices for the development and management of AI. 

In January, I shared my thoughts on potential outcomes of the AI Impact Summit. Some of these predictions came to fruition. However, my thought that this year's AI Summit would serve as a bellwether for the tone for the rest of the year appears to be less certain. 

While the AI Impact Summit was an important event with a number of new announcements and financial commitments, there was not a standout moment, like U.S. Vice President JD Vance's infamous speech at last year's AI Action Summit in Paris, that will shift the direction of AI policy this year. 

General sentiment and outcomes

This is not to say this year's summit was less impactful. Indeed, the event validated many sentiments that bubbled to the surface in late 2025 and early this year. Some of these themes, including an emphasis on increasing AI adoption, speculation on labor market impacts and sector-specific AI application, were key topics at the conference. However, there was also much more emphasis on global cooperation and the need for shared tooling than I anticipated. It almost seemed like a shift back to the discussions that happened at the original AI Safety Summit in the U.K. 

If you're in the AI community, your LinkedIn feed was probably filled with posts about the summit. While the posts contain a lot of what you would expect, I shared a similar reaction to Divij Joshi, research fellow at ODI Global. Like him, I was hopeful that this summit would become a forum for advancing genuine AI governance efforts with civil society actors at the table. It was also a pleasant surprise to hear that responsible AI is considered an important aspect of these governance efforts. 

Interestingly, many of these sentiments and outcomes were captured in the AI Impact Declaration that was endorsed by 91 countries. The declaration starts with clarifying one of the summit's principles, "Welfare for all, Happiness of all," which set the tone for the non-binding agreement that recognizes people are at the heart of building and deploying AI, and as a result, have agency over its outcome. At the Global Partnership for AI Council meeting that took place alongside the summit, Canada's Minister of Artificial Intelligence and Digital Innovation Evan Solomon acknowledged, "AI will define the era, people will define the impact." 

During that meeting, the strong majority of attendees discussed the importance of global cooperation in sharing resources and expertise to ensure the value of AI is maximized. While this might seem like typical high-level governmental language, the emphasis on cooperation was a shift in tone from 2025. Even with continued emphasis on sovereignty and regional realities, the broad recognition that partnership can advance national interests was striking.  

Language

In addition to these higher-level outcomes, there were more tangible themes that emerged. The discussion of language, more specifically languages, came up in almost every session I participated in or attended. To scale AI adoption in the way that many envision, it is important to understand that, to date, much of the work in AI has been primarily conducted in English. With less than 20% of the world's population identifying as English speakers, there is a clear need for AI to be trained and deployed in many other languages in order to maximize adoption. 

To address this challenge, India's government announced eight indigenous foundation AI models engineered for India's linguistic complexity. These models are set to provide generative AI capabilities for all of India's 22 recognized languages. 

The Global South

As anticipated, one of the most significant takeaways was the fact that this was the first AI Summit held in the Global South. Having been on the ground and spoken with locals from across this region, it is clear that the realities faced in the Global South differ significantly from the perspectives that have shaped the AI development and governance community, which remains largely rooted in Western experiences and resources.  

The Global South has different realities than the North, including human rights, access to resources, infrastructure and a multitude of languages, as discussed above. And within the Global South, each country, region, and city have different priorities and challenges that will need to be addressed when focusing on adoption. 

One example that I really appreciated from an AI governance and finance panel about India was that the country remains a conversational market. However, this has resulted in lower online commerce than other nations by approximately 30%. The speaker raised the point that agentic AI could be a significant change for India so that there is still the opportunity to have more conversation through the online purchasing process. 

And for those who aren't familiar with Delhi traffic, my big question is how automated vehicles are going to survive there. They will certainly require a different training set than San Francisco!  

IAPP on the ground across India

In addition to participating in the AI Impact Summit, we also had a delegation tour to meet the IAPP community in Bangalore, Mumbai and Delhi. This was a phenomenal opportunity to meet digital governance practitioners and learn more about their work. The energy, enthusiasm and ideas coming out of India were inspirational. The curiosity to better understand AI and its impact was palpable. 

We hosted local community events in each city with the objective of understanding the evolving role of digital governance professionals and the impact they have on increasing the adoption of digital technologies including AI. With both local and global speakers, it was interesting to see some of the contextual realities I mention above identified while also seeing how similar governance practices can be independent of what country or industry you are operating in. 

One key takeaway from these meetings was the experimentation by some of the participants with using AI more extensively in governance practices. The concept of using AI to govern AI did come up in many conversations throughout the tour. We will continue to follow how this practice is working and how it is being implemented in different organizations. 

We were pleased to present two panels at this year's summit, "Implementing AI Standards for Global Prosperity in an Era of Agentic AI," and "How Trust and Safety Drive Innovation and Sustainable Growth."

Next steps

As anticipated, Switzerland announced that they will host the 2027 AI Summit. No date has been announced yet, but there is speculation the event will occur in July instead of February as the last two have been. Ireland and Saudi Arabia also announced that they will host Global AI Summits later this year. It is expected that these will be more regional in focus. 

The IAPP will continue to share more insight into these increasingly important global forums for AI and its governance.