The New Delhi Declaration: What 88 Signatories Agreed to at the AI Impact Summit

On 19 February 2026, the AI Impact Summit in New Delhi concluded with the formal adoption of the New Delhi Declaration by 88 countries and international organisations, according to Outlook Business. The summit brought together more than 20 heads of state, 60 ministers, and over 500 global AI leaders to discuss the trajectory of artificial intelligence development. While the summit established a high-level consensus on the principle that the benefits of AI should be shared by humanity, it also revealed a significant divide regarding how such principles should be enforced through international law.
The Principle of Shared Benefits and the Enforcement Gap
The central tenet of the New Delhi Declaration is the commitment by signatories to the principle that the promise of AI is best realised when its benefits are shared by all of humanity. This principle aims to address concerns regarding the digital divide and the potential for AI technologies to exacerbate global inequality. However, a critical observation from the summit is the absence of any binding enforcement mechanisms within the declaration. As noted by techUK, the document remains a statement of intent rather than a regulatory framework with teeth.
This lack of enforcement creates a dual-track reality for international AI policy. On one hand, the declaration provides a diplomatic baseline for cooperation, signaling that 88 entities are aligned on the humanitarian goal of equitable AI distribution. On the other hand, the lack of oversight means the actual implementation of these "shared benefits" remains entirely at the discretion of individual nation-states and corporations. One reading is that the summit prioritized broad consensus over meaningful accountability, a common trade-off in multilateral diplomacy when navigating sensitive technological sovereignty issues.
The absence of binding enforcement mechanisms suggests a focus on voluntary cooperation rather than regulatory unification, leaving the actual delivery of shared AI benefits to the whims of national policy and private sector investment.
Why It Matters
Geopolitical Divergence on Global Governance
The summit highlighted a profound tension between the desire for collective international standards and the preservation of national technological sovereignty. While the declaration was endorsed by a large majority of attendees, the United States delegation maintained a clear and explicit position. Specifically, White House official Michael Kratsios stated that the United States totally rejects any form of global governance of AI, according to techUK. This position represents a significant hurdle for any attempt to create a centralized, UN-style regulatory body for artificial intelligence.
This divergence creates several plausible scenarios for the future of AI regulation. In one scenario, the world moves toward a fragmented regulatory terrain, where "pockets" of high-regulation zones (such as the EU) exist alongside "innovation-first" zones (such as the US). In an alternative scenario, the principle of "shared benefits" expressed in the New Delhi Declaration serves as a soft-power mechanism, eventually influencing global norms through economic and diplomatic pressure rather than through formal legal structures. The rejection of governance by a major power like the US suggests that the "New Delhi model" is one of voluntary alignment rather than mandatory compliance.
Economic Commitments and Infrastructure Scaling
Despite the lack of a formal governance structure, the summit produced substantial material commitments. There are reported commitments totaling over $250 billion aimed at infrastructure-related development, according to techuk.org. This capital is intended to support the hardware and energy requirements necessary to deploy AI at a global scale. The scale of this investment underscores the fact that while the legal frameworks remain contested, the physical and economic foundations of the AI economy are being aggressively established.
The distinction between "infrastructure commitments" and "direct investment" is a critical nuance. Infrastructure-related spending often involves large-scale capital expenditure on data centers, subsea cables, and power grids, which may be driven by both public policy and private-sector necessity. For investors, this indicates that the most immediate growth may lie in the physical layer of AI deployment, regardless of whether the software layer is governed by international treaties.
| Commitment/Entity | Value or Scale | Primary Source Context |
|---|---|---|
| Global Infrastructure Commitments | Over $250 billion | techUK reports total scale |
| IndiaAI Mission | ₹10,000 crore ($1.2bn) | techUK reports financial commitment |
| Indian Student Pledges | 250,946 students | Outlook Business reports record |
National Strategies and the Role of Public Mobilization
The summit also served as a platform for showcasing domestic AI initiatives that serve as blueprints for other nations. India's AI strategy was particularly visible, both through state policy and public engagement. The IndiaAI Mission, which received an approval of ₹10,000 crore ($1.2bn) in March 2024, serves as a cornerstone of the country's approach to building indigenous capability. This mission is designed to bolster the local ecosystem through massive financial input into compute resources and research.
Beyond government funding, the summit highlighted the role of mass mobilization in establishing "responsible AI" norms. In a notable display of public engagement, 250,946 students in India pledged to practice responsible AI within a 24-hour period, according to Outlook Business. This level of participation suggests that nations are increasingly looking toward civil society and education to build the social license required for AI integration, even when formal international governance remains elusive.
Implications for Global Stakeholders
The outcome of the New Delhi Summit provides distinct signals for various actors in the technology and political economy. For policymakers, the key takeaway is the realization that the era of a single, unified global AI regulator may be stalled by the preferences of major technological powers. Instead, they should prepare for a terrain defined by bilateral and multilateral "norms-based" cooperation rather than "law-based" enforcement.
Investors and executives should monitor the following indicators to assess the actual impact of the New Delhi Declaration:
- The flow of the $250 billion in pledged infrastructure funds into specific, verifiable projects in developing nations.
- The development of industry-led standards that may emerge as a practical substitute for formal government governance.
- The emergence of "AI-free trade zones" or "high-regulation corridors" that could impact the movement of data and compute across borders.
- The specific metrics that signatories use to define "shared benefits" in their subsequent national reports.
In my analysis, the New Delhi Declaration represents the successful creation of a global "moral consensus" that lacks a "legal enforcement" mechanism. This creates a significant opportunity for those who can define the technical and economic standards that will govern AI, as the vacuum left by the absence of global governance will likely be filled by technical interoperability and economic necessity rather than international law.
Sources: negd.gov.in, outlookbusiness.com, techuk.org.