Ms. Khyati Singh, Research Analyst, spoke on “AI Impact Summit 2026: Key Takeaways and India’s Strategy” at the Monday Morning Meeting held on 16 March, 2026. The Session was moderated by Col. Nestor Derrick D’Souza, Research Fellow. Ambassador Sujan R. Chinoy, Director General, MP-IDSA and the scholars of the Institute attended the discussion.

Executive Summary

The Speaker traced global AI governance from Bletchley Park to the AI Impact Summit 2026, where India advanced the MANAV framework of moral, accountable, sovereign, accessible and valid AI, alongside initiatives for democratic diffusion and indigenous language models. It then examined AI’s weaponisation in U.S. operations in Venezuela and Iran, showing its role in autonomous warfare and exposing the risks of dependence on commercial models. The Speaker highlighted the duality in AI-driven technologies, while forwarding the central argument that India must pursue sovereign, balanced AI development.

Detailed Report

Col. D’Souza began his Opening Remarks by asserting that human evolution has progressed through distinct ‘ages’, from agriculture to industrialisation, then the information technology or ‘knowledge’ age, and now the era of Artificial Intelligence (AI) . Today, AI shapes everyday life, solving routine and complex problems alike, even as individuals increasingly trade privacy for convenience under constant digital surveillance. Global leaders emphasise that dominance in AI will determine future world power, intensifying international competition. In this context, India’s AI Impact Summit highlighted a shift from dialogue to demonstrable impact, guided by people, planet, and progress, and inspired by Vasudeva Kutumbakam. The Summit addressed key concerns such as child safety, job disruption, ethics, environmental impact, and equitable growth, while also showcasing India’s AI capabilities and fostering collaborative policy development through broad stakeholder engagement.

Ms. Khyati Singh began her presentation by stating that the trajectory of global AI governance can be understood through a series of international summits beginning in 2023. The first, held in the United Kingdom at the Bletchley Park AI Safety Summit, marked a defining moment as countries collectively came together to deliberate on AI. Attended by around 100 representatives from 30 technologically advanced nations, it focused narrowly on AI safety, risk mitigation, and concerns surrounding artificial general intelligence escaping human control. The primary outcome was the establishment of AI Safety Institutes aimed at standardising testing methodologies before commercial deployment.

Ms. Khyati also pointed out that this safety-centric approach continued into the AI Seoul Summit of May 2024, where efforts were made to formalise concerns through frontier AI safety commitments. Sixteen leading technology companies agreed to assess risk across the AI lifecycle, define intolerable thresholds, and halt developments that could significantly harm humanity. However, compliance remained uneven, with several corporate signatories establishing internal safeguards inconsistently.

The Speaker further highlighted that attempts to build geopolitical consensus during the Paris Summit remained incomplete. Countries such as the United Kingdom and the United States refrained from endorsing binding ethical standards, arguing that such frameworks could hinder AI development, particularly in contexts such as weapon development. This reluctance reflected a broader tension between ethical AI and strategic use, especially as AI-enabled weapons and assistant systems have already been deployed in recent conflicts.

The AI Impact Summit in India in 2026 marked a definitive geographical departure, being the first such event hosted in the Global South. Held at Bharat Mandapam in New Delhi from February 16 to 21, it witnessed participation of over 600,000 individuals in person and approximately 900,000 virtually. India was projected as a leader for both the AI world and the Global South, signalling a shift in the centre of gravity of AI discourse.

According to Ms. Khyati the most important outcome of the Summit was the articulation of a human-centric governance framework termed MANAV. The acronym encapsulates a unified structure: Moral and Ethical Systems emphasising fairness, transparency, and continuous human oversight embedded directly within algorithmic architecture; Accountable Governance ensuring transparent administrative rules and robust oversight mechanisms, supported by the India AI Mission to democratise datasets and mandate safe AI principles; National Sovereignty asserting the inalienable right of States to protect domestic data within their own cultural, linguistic, and legal frameworks, guided by the principle “whose data, his right”; Accessible and Inclusive AI demanding systems that function across low-resource environments and diverse linguistic landscapes beyond English-speaking populations; and Valid and Legitimate Systems focusing on verifiability and trustworthiness of outputs to counter deepfakes and AI hallucinations through filtered, legitimate data sources.

This comprehensive framework positions India as an ideological vanguard of the Global South, advocating a balanced approach that avoids colonial-style data extraction and instead promotes a Make in India model, allowing countries to store data within their own boundaries and build local cloud and dataset infrastructures. The Summit’s diplomatic outcomes included the New Delhi Declaration on AI Impact, endorsed by approximately 92 countries and international organisations, including major strategic rivals such as the United States, China, Russia, and the United Kingdom, reflecting a broad global consensus.

Another significant outcome was the Charter for Democratic Diffusion of AI, supported by 22 countries and multiple institutions. This voluntary framework aims to break the oligopolistic control of foundational AI models by promoting local innovation ecosystems, technology transfer, and resilient infrastructure while respecting sovereign goals. Supporting initiatives included the Global AI Impact Commons to scale and share use cases, the Trusted AI Commons as a repository of safety benchmarks and best practices, and an equitable AI transition playbook developed with the International Labour Organization to address labour market disruption.

The Summit also highlighted India’s breakthrough use case models. Mission Bhashini and VoicERA were presented as national language infrastructure integrating over 350 AI models across 22 official languages, enabling even illiterate individuals to access AI services through voice by simply dialling a number. These models demonstrated how AI can be made accessible to nearly 130 crore people in their own linguistic contexts.

Sarvam AI was introduced as a domestic entity developing sovereign foundational models trained on Indian datasets, optimised for regional scripts and linguistic nuances. Its models enable real-time translation and application across sectors such as film production, reducing costs and expanding reach. BharatGen AI, a strictly government-funded multimodal model, integrates text, speech, and document vision capabilities, ensuring secure and sovereign public service delivery without third-party risk. Inya VoiceOS, a multilingual speech model, operates effectively under severe low-bandwidth conditions, ensuring accessibility even in regions with limited connectivity.

The discussion then shifted to the global scale, where alongside democratic and inclusive AI, the weaponisation of AI has become evident. In January 2026, the United States launched Operation Absolute Resolve in Venezuela, capturing President Maduro within hours through AI-driven cyber and intelligence systems. AI enabled manipulation of grid systems, real-time tracking of locations, and predictive scenario modelling, demonstrating effects-based operations with high precision and minimal human damage.

This was followed by Operation Epic Fury in Iran, showcasing autonomous kinetic saturation through large-scale deployment of low-cost uncrewed combat systems. AI-enabled drone swarms communicated autonomously, adjusted flight paths, and executed coordinated multi-vector strikes while overwhelming defence systems. Simultaneously, AI-assisted defence systems such as Iron Dome and Patriot batteries intercepted retaliatory attacks, illustrating AI’s dual role in offensive and defensive operations.

These operations exposed structural vulnerabilities, particularly the dependence on commercial AI platforms. The use of external models such as Claude led to disputes between private companies and the U.S. Government, as corporate entities objected to military usage that violated their guidelines. Despite official discontinuation announcements, such models continued to be used due to deep integration, highlighting the difficulty of replacing embedded AI systems without equivalent substitutes.

In her concluding remarks, Ms. Khyati opined that the key takeaway for India is the need to maintain a dual AI strategy, i.e., promoting development, security, and safety while asserting sovereignty. Ethical considerations must coexist with strategic imperatives through a balanced approach. India must develop indigenous, government-funded AI models with locally stored datasets to avoid dependence on private or foreign entities. At the same time, it must handhold small companies and emerging players to achieve scale, ensuring that domestic innovation ecosystems can match global advancements while remaining rooted within national systems.

Q&A Session

In the discussion that followed, Ambassador Sujan R. Chinoy, Director General, MP-IDSA, noted that the rapid advancement of AI will soon surpass conventional research and analysis, making its integration essential for scholars. He emphasised that India must balance concerns over job losses with strengthening capabilities in data sovereignty, models, software, and computing power. Instead of pursuing perfect systems, the priority should be early and practical diffusion. He highlighted the need to use AI to expand access to science and technology beyond English-speaking elites and across defence, business, and daily use.

The discussion also covered AI in warfare, ethical guardrails, defence applications, the role of private entities, human agency, and challenges of data and indigenisation in India.

Report prepared by Ms. Meghna Pradhan, Research Analyst, Centre for North America and Strategic Technologies, New Delhi.