Leaders’ Plenary | Global Vision for AI Impact and Governance l AI Impact Summit 2026

Contents

Executive Summary

The AI Impact Summit 2026, hosted in New Delhi by India, brought together global leaders, ministers, and international organizations to reframe AI governance around inclusion, equity, and human-centered development. Rather than treating AI as a technology race driven by speed and scale, the summit emphasized AI's role in serving people, particularly in the Global South, with the overarching principle of "Vasudhaiva Kutumbakam" (the world is one family) and the commitment to "welfare for all, happiness for all." The summit produced the Delhi Declaration and identified seven key pillars (represented as "chakras") for responsible AI governance.

Key Takeaways

AI is not a race—it's a transformation requiring collective responsibility. Speed and scale alone do not define success; the measure is whether AI serves all humanity, particularly the most vulnerable, and strengthens democratic institutions rather than eroding them.
The Global South must lead, not follow. This summit—the first on AI hosted in a Global South country—signaled that AI governance decisions cannot be made by wealthy nations and tech companies alone. India's model of inclusive, multilingual, and human-centered AI development offers an alternative to narratives of technological concentration.
Infrastructure investment in developing nations is not charity—it is essential to global stability and equity. Without access to compute, data, and talent, billions of people will be locked out of AI benefits, and countries will lose agency in shaping global standards.
Trust is built through transparency, participation, and demonstrated benefit. Governance frameworks, regulations, and standards mean little without institutional credibility and visible improvements in citizens' lives.
Language and culture are not peripheral to AI—they are fundamental. AI systems designed only for wealthy English-speaking populations will inherently exclude billions and reinforce existing power asymmetries.

Key Topics Covered

AI Governance & International Frameworks: Global dialogue mechanisms, UN's new scientific panel on AI, multilateral cooperation vs. fragmented approaches
Digital Sovereignty & Infrastructure: Computing power concentration, data center development, energy requirements for AI, sovereign AI capabilities
Global South Representation: Bridging the AI divide, technology access disparities, capacity-building in developing nations, Global South leadership in AI policy
AI Ethics & Human Rights: Transparency, accountability, bias mitigation, protection of minors, informed consent, cultural context in AI
Economic Impact & Labor: Job displacement concerns, skills development, worker reskilling, productivity gains, wage inequality
Environmental Sustainability: Green energy for data centers, carbon footprint of AI infrastructure, climate resilience through AI
Language & Cultural Inclusion: Multilingual AI models, representation of non-English languages, indigenous language preservation
Education & Capacity Building: AI literacy, upskilling programs, talent pipelines, public-private partnerships
Regulatory & Trust Frameworks: Risk-based regulation, national AI strategies, regulatory sandboxes, standards development
Regional & Bilateral Initiatives: India-Switzerland partnership for 2027 summit, India-EU cooperation, regional data centers, South-South collaboration

Key Points & Insights

AI as Infrastructure & Power: Multiple speakers framed AI not merely as a tool but as critical national infrastructure that determines geopolitical influence, economic competitiveness, and sovereignty. Computing power, data, and algorithmic control are now measures of state power comparable to territorial control.
Concentration of Power Risk: Small number of technology companies and wealthy nations control foundational AI resources (compute, large language models, datasets). Without deliberate redistribution, AI risks creating a new form of inequality—an "AI divide"—that will perpetuate and amplify existing global disparities.
Wisdom Must Guide Intelligence: Prime Minister Modi and others emphasized the Sanskrit distinction between aparavidya (technical knowledge) and paravidya (wisdom). Technical capability without ethical wisdom risks deepening inequality, enabling misinformation, and accelerating beyond governance capacity.
AI Dividend Must Be Shared: Technological revolutions create wealth, but distribution is never automatic. The summit stressed that tangible benefits—improved healthcare, education, job creation—must reach farmers, nurses, teachers, and small entrepreneurs, not just tech companies and shareholders.
Democratic Participation in AI Governance: AI governance cannot be decided by technologists or a handful of nations alone. Inclusive multilateral frameworks (through the UN, regional bodies, civil society participation) are essential to legitimacy and effective implementation.
Infrastructure Gaps Require Action: Developing nations lack affordable compute access, representative datasets, skilled talent pipelines, and regulatory capacity. The summit called for practical bilateral and multilateral mechanisms—not just capacity-building promises—to close these gaps.
Skills & Education as Foundation: Education must lead AI strategy. Investments in reskilling workers, digital literacy, and talent development are as critical as infrastructure investment. Without human capacity, technology remains inaccessible.
Trust as Competitive Advantage: Small nations cannot compete on scale or capital, but can compete on trust, transparency, and values-based governance. Estonia, Slovakia, Switzerland, and others positioned themselves as trustworthy AI partners and testing grounds for responsible innovation.
Multilingualism & Cultural Context: AI systems that do not reflect linguistic and cultural diversity will fail to serve populations equitably. Low-resource languages and locally relevant models must be prioritized to ensure AI is "locally relevant, culturally grounded, and socially legitimate."
Job Transformation, Not Purely Automation: IMF and ILO research shows AI will transform more jobs than eliminate them, but middle-skill, routine tasks face compression. Policy must anticipate displacement, invest in transition support, and ensure innovation advances human dignity and rights.

Notable Quotes or Statements

Prime Minister Narendra Modi: "Whether AI benefits humanity or not will depend entirely on paravidya [wisdom]. So perhaps the most important question of our time is not how intelligent our machines will become but whether we will remain wise enough to guide them."
Prime Minister Modi (on AI as infrastructure): "Technology moves quickly. It is measured in years or even in months. Institutional trust moves more slowly. It is built over generations. AI will test our ability to align these very different rhythms."
Prime Minister Krista Kaljulaid (Estonia): "Small countries are unable to compete with large ones in terms of capital and computing power. We can however compete on trust, transparency and values based governance."
Prime Minister Robert Fico (Slovakia): "The world does not need more words. It needs results."
UN High Commissioner for Human Rights, Peggy Hicks: "We need to narrow the gap between those at the forefront of digital technologies and those most likely to benefit from them."
IMF official: "We have to pay more attention to how we prepare people for the job market of tomorrow... AI is coming like a tsunami hitting the job markets."
PM of Greece: "A world in which technology is weaponized to coerce trusted partners or where excessive regulation becomes a tool to suppress innovation is a world where collective innovation declines."
Philippine delegate (implied): "If technology is not sent or channeled down to those who actually need it, we may find ourselves in a world where we have solutions but ultimately the problems that the solutions were created for still remain."
ITU Secretary-General, Dorin Bogdan Martin: "The true measure of success will be whether it reaches everyone in every country and in every local community... We cannot let the digital divide become the AI divide."

Speakers & Organizations Mentioned

Government Leaders & Ministers

Prime Minister Narendra Modi (India) – Host, keynote speaker
Kaja Kallas (Estonia) – Prime Minister
Aleksandar Vučić (Serbia) – President
Robert Fico (Slovakia) – Prime Minister
Ignazio Cassis (Switzerland) – Foreign Minister
Alain Berset (Liechtenstein) – Hereditary Prince
Jairam Ramesh (India) – Minister of Environment, Forest and Climate Change
Lotay Tshering (Bhutan) – Foreign Minister
Nayib Bukele (El Salvador) – President
Mitsotakis (Greece) – Prime Minister
Petteri Orpo (Finland) – Prime Minister
Giorgia Meloni (Italy) – Prime Minister
Pedro Sánchez (Spain) – Prime Minister
Hassan Rouhani (Iran) – Referenced but not present
Abdulla Shaheen (UAE)
Kagame (Rwanda)
Cyril Ramaphosa (South Africa)

International Organization Leaders

António Guterres – UN Secretary-General
Kristalina Georgieva – IMF Managing Director
Gilbert F. Houngbo – ILO Director-General
Dorin Bogdan Martin – ITU Secretary-General
Peggy Hicks – UN High Commissioner for Human Rights
Ibrahim Thiaw – UN Environment Programme Executive Director (referenced)

Ministers & Government Officials

Ashwini Vaishnaw (India) – Minister of Railways, Information Technology
Kang Kyung-wha (South Korea) – Deputy PM, Minister of Science and ICT
Ebba Busch (Sweden) – Deputy PM, Minister of Energy
Josephine Teo (Singapore) – Minister
Raymon Soders (Latvia) – Minister
Eduardas Vaitkus (Lithuania) – Minister of Economy
Vibhuti Bhatnagar (Malaysia) – Minister of Digital
Makshut Shahadev (Russia) – Minister of Digital Development
Amal al-Fala Suchniey (Morocco) – Minister Delegate
Dr. Shen (New Zealand) – Minister for Science, Innovation
Siham Al-Mamari (Oman) – Minister of Transport
Karina Olsen (Norway) – Minister of Digitalization
Gonçalo Matias (Portugal) – Minister to PM
Chris Muimo (Uganda) – Official representative
Dingamoji Futy (Zimbabwe) – Deputy Minister
William Kabogo (Kenya) – Cabinet Secretary
Abdullah bin Saraf al-Ghamdi (Saudi Arabia) – President, Saudi Data and AI Authority
Sheri Kabir (Tajikistan) – Minister of Industry
Angela Kaiuki (Tanzania) – Minister of Communications

International Institutions

United Nations (UN), including UN General Assembly, UN High Commissioner for Human Rights office
International Monetary Fund (IMF)
International Labour Organization (ILO)
International Telecommunication Union (ITU)
International Committee of the Red Cross (ICRC)
World Meteorological Organization (WMO)
World Intellectual Property Organization (WIPO)
OECD
European Union (EU)
APEC (Asia-Pacific Economic Cooperation)

Companies & Tech Platforms Mentioned

Microsoft, Google, AWS – Major cloud/AI providers
NVIDIA – GPU chip manufacturer
Tata Power, Adani Power – Indian energy companies
InfoBep, Reheats, Microlink, GDON, Infinum – Croatian tech companies

Technical Concepts & Resources

AI Infrastructure & Computing

GPU chips (especially NVIDIA generation)
Data centers and sovereign compute infrastructure
Supercomputing clusters (e.g., Lumi in Finland, Perón in Slovakia)
High-performance computing (HPC)
Cloud infrastructure and sovereign cloud
AI factories / AI gigafactories (e.g., Slovakia's project, European initiatives)
Computing power measured in gawatts (energy requirement metric)
Edge AI and on-device models

AI Models & Language Models

Large Language Models (LLMs) – general reference
Foundational models – base models used to build applications
ALAM (Saudi Arabia's Arabic Large Language Model) – ranked second globally in Arabic LLM performance
Moin AI (Oman's national LLM trained on local culture, policies, traditions)
Multilingual models – explicitly prioritized for global inclusion
Agentic AI – autonomous systems capable of independent action

Data & Governance

Data lakes – centralized repositories (e.g., Lithuania's national data lake)
Data factories – infrastructure for curating and preparing datasets
Data protection regulations – GDPR analogues, national frameworks
Data sharing frameworks (e.g., Malaysia's Data Sharing Act of 2025)
Open data initiatives – public datasets for research and innovation
Regulatory sandboxes – controlled environments for testing compliant AI solutions

AI Applications Across Sectors

Healthcare AI: diagnostic support, CT/MRI analysis, predictive analytics, preventive care
Education AI: adaptive learning, personalized learning paths, teacher support
Agricultural AI: crop optimization, resource management
Climate & Environment AI: weather prediction, climate resilience modeling
Public services AI: government efficiency, service delivery, case management
Financial AI: fraud detection, credit assessment
Security & Defense AI: autonomous systems (with caveats about military AI governance)
Image recognition and computer vision
Autonomous systems (robots, vehicles) – with emphasis on human oversight

Safety, Ethics & Governance Frameworks

AI safety research – formal discipline, priorities set by scientific consensus
Explainability/Interpretability – understanding AI decision-making
Bias detection and mitigation
Transparency requirements – clear disclosure of AI use and limitations
Human-in-the-loop systems – mandatory human oversight for critical decisions
Algorithmic accountability – mechanisms to contest AI decisions
Risk-based regulation – tailored safeguards based on harm potential
AI act (EU) – mentioned as a reference model
Ethical AI frameworks – human rights, fairness, accountability alignment
AI governance panel (UN's new scientific panel on AI)

International & Policy Frameworks

Global Digital Compact (UN initiative)
AI Impact Summit framework – this summit's outputs
Delhi Declaration – summit's key agreed statement
Seven Chakras – India's framing of AI governance pillars (metaphorical framework)
REAIM summit – international dialogue on responsible AI in military domain
National AI strategies – country-level roadmaps (Zimbabwe, Slovakia, Malaysia, etc.)
Public-Private Partnerships (PPPs) – collaborative governance models

Measurement & Metrics

Teraflops – computing performance measure (explicitly rejected in favor of human impact metrics)
GDP contribution from AI (e.g., Tajikistan targeting 5% by 2040)
Digital literacy rates
Digital government rankings (OECD Index; Portugal ranks 3rd globally)
AI readiness assessments (e.g., Zimbabwe's methodology)
Job displacement rates and labor market impact studies

Notable Projects & Initiatives

UPI (Unified Payments Interface) – India's digital payment system, cited as example of scalable digital infrastructure
Aadhaar – India's biometric identification system (1.4 billion people)
India Stack – foundational digital public infrastructure model
AI Leap initiative (Estonia) – public-private partnership for AI education
ASD.AI program (Estonia) – national AI implementation initiative
Galufu Mindfulness City (Bhutan) – sustainable AI innovation hub
Bratislava AI Forum (Slovakia, November 2025) – regional dialogue
Laureng AI (Singapore) – AI community platform
AI Park (Singapore, proposed)
LitAI (Lithuania) – national AI factory
SEMEI (Saudi Arabia) – 1 million citizens to AI training program
Hexagon (Saudi Arabia) – world's largest governmental data center, launched December 2025
Area AI (Tajikistan) – free zone for AI investment
Tanzania Technology Stack – digital public infrastructure development
Transaspian Fiber Optic Line (Kazakhstan) – critical data infrastructure

Research & Collaboration Networks

ELLIS Institute (Europe) – AI research
UN's new scientific panel on AI – established to guide global governance
AI for Good initiatives (ITU, others) – solving humanitarian challenges
Grand challenges in AI for science (Singapore) – material science, life science, agriculture, computer science
International Scientific Exchange on AI Safety (Singapore hosts 2nd edition, 2026)
Singapore Consensus – global AI safety research priorities

Context & Significance

This summit represents a historic shift in AI governance narrative:

**First AI governance summit hosted in the Global