Global AI Accord Reached, But Implementation Hurdles Remain
Latest Delhi hosted the India AI Impact Summit this February, culminating in the New Delhi Declaration – a landmark agreement endorsed by 86 countries and two international organizations. The declaration aims to foster a globally inclusive, human-centric and development-oriented AI framework. However, the path from consensus to concrete action is fraught with challenges, as highlighted during the summit’s discussions.
The Promise of a Human-Centric AI
India’s Minister for Electronics and IT, Ashwini Vaishnaw, emphasized the declaration as a demonstration of “broad global support” for a vision prioritizing AI access, skilling, and responsible deployment. The agreement seeks to expand AI benefits to developing economies, strengthen its application in crucial sectors like healthcare and education, and build trust through accountability.
The declaration isn’t merely aspirational. It’s intended to guide national policies, encourage cross-border research, and establish consistent, measurable standards for AI development and implementation. A key focus is addressing pressing challenges like algorithmic bias, cybersecurity risks, potential workforce disruption, and broader societal impacts.
Unequal Access: A Challenge for the Global South
Despite the broad endorsement, concerns remain about equitable access to AI’s benefits. Leaders from smaller and developing nations voiced concerns during the summit, noting significant limitations in financing, research and development capabilities, and access to capital. Without external partnerships, these nations struggle to invest in the necessary infrastructure and expertise.
Prime Minister of Mauritius, Navin Ramgoolam, and Vice-President of Seychelles, Sebastien Pillay, underscored the need for technology transfer, legal readiness, and sustained cooperation to truly democratize AI for vulnerable economies. The declaration acknowledges this need, calling for infrastructure, finance, and technical cooperation.
The US Position: Sovereignty Over Governance
A contrasting perspective emerged from the United States. Michael Kratsios, Director of the White House Office of Science and Technology Policy, signaled resistance to centralized global oversight, advocating for “sovereign AI capability.” The US approach prioritizes trade and partnerships over supranational regulatory structures, positioning American AI as “open for business.”
This divergence highlights a fundamental tension: balancing the desire for global cooperation with national interests and strategic advantages. The declaration, while articulating shared commitments, doesn’t resolve this tension, leaving implementation contested.
AI as Infrastructure: A Geopolitical Shift
The summit also highlighted the growing recognition of AI as critical infrastructure. Serbian President Aleksandar Vučić argued that control over data, algorithms, and domestic expertise is essential for national sovereignty in the 21st century. This perspective underscores the geopolitical implications of AI, as control over these elements translates to power and influence.
The consolidation of large language models, compute clusters, and semiconductor supply chains within a limited number of economies and corporations further exacerbates this issue. The question of who owns the infrastructure and defines the standards is central to achieving global equity in AI.
Compute Power, Energy, and Human Oversight
Slovakian President Peter Pellegrini emphasized the importance of access to skills, tools, and fair conditions for innovation, stating that AI “must not stay in the hands of few.” He highlighted computing power as the “new infrastructure,” alongside energy-intensive data centers, advanced GPUs, and stable legal frameworks.
A recurring theme throughout the summit was the necessity of human oversight. Maintaining human control and responsibility was consistently reinforced as a baseline principle for responsible AI development and deployment.
What the Declaration Achieves – and Doesn’t
The New Delhi Declaration represents a significant step towards global consensus on responsible AI. It articulates shared commitments to inclusivity, access, and cooperation. However, it’s a non-binding agreement lacking enforcement mechanisms. It doesn’t directly address the concentration of compute power or reconcile differing philosophies on global governance.
the declaration codifies principles but leaves the complex task of implementation contested, reflecting the inherent challenges of multilateral agreements in a rapidly evolving technological landscape.
FAQ: The New Delhi Declaration and the Future of AI
Q: Is the New Delhi Declaration legally binding?
A: No, the declaration is non-binding. It serves as a statement of intent and a framework for future collaboration.
Q: What are the key priorities outlined in the declaration?
A: The declaration prioritizes expanding AI access for developing countries, strengthening digital public infrastructure, and ensuring trust through transparency and safeguards.
Q: What is the US stance on global AI governance?
A: The US rejects centralized global oversight, advocating for “sovereign AI capability” and prioritizing trade and partnerships.
Q: What role does infrastructure play in the AI landscape?
A: AI is increasingly viewed as critical infrastructure, with control over data, algorithms, and compute power becoming central to national sovereignty.
Did you know? The India AI Impact Summit 2026 attracted 20 heads of state and delegates from over 80 countries.
Pro Tip: Stay informed about national AI strategies and policies to understand how your country is responding to the challenges and opportunities presented by AI.
What are your thoughts on the New Delhi Declaration? Share your perspective in the comments below!
