The objective of the Delhi summit must be to lay down foundational ground rules for the AI era
“If the Internet was transformative, artificial intelligence (AI) is transmogrifying. The critical difference between the dotcom boom and the AI explosion is the substrate upon which it is built. The Internet required a quarter-century to achieve its current ubiquity, a period of gradual, albeit rapid, adoption and infrastructure development. AI, in stark contrast, is being grafted onto the existing, globalized digital nervous system that the Internet created.”
The Internet was the most audacious experiment in anarchy, and it has succeeded beyond the wildest imagination of its initial progenitor, the US Defense Advanced Research Projects Agency (DARPA). It represents the largest ungoverned space on earth. Never before in the history of humankind has so much power concurrently resided on 5.5 billion fingertips around the world.
There is more data that is churned out daily than from the dawn of civilization to the turn of the third millennium. The future of humankind lies on the intersection of a brick-and-mortar civilization that evolved over an eternal length of time and a virtual civilization that is still metamorphosing.
The Internet that came in vogue in the mid-1990s was truly transformative in the way it has shaped life, leisure and work patterns. However, it has also been weaponized by state, semi-state and other nefarious characters, including people who have a grievance as to why they were born in the first place.
The Dark Web or the Deep Net is a classical manifestation of the noxious underbelly of the Internet. However, despite a host of acronyms that ostensibly govern the Internet — Internet Governance Forum (IGF), Internet Architecture Board (IAB), Internet Corporation for Assigned Names and Numbers (ICANN), Internet Engineering Task Force (IETF) and the World Wide Web Consortium (WSC), to name a few — there are no agreed rules of engagement in this virtual civilization.
The First Amendment absolutists of the Silicon Valley and other innovation hubs did not allow a cohesive global governance structure for the Internet to get institutionalized. This has led to myriad challenges, including the balkanization of the Internet into spheres of influence.
If the Internet was transformative, artificial intelligence (AI) is transmogrifying. The critical difference between the dotcom boom and the AI explosion is the substrate upon which it is built. The Internet required a quarter-century to achieve its current ubiquity, a period of gradual, albeit rapid, adoption and infrastructure development. AI, in stark contrast, is being grafted onto the existing, globalized digital nervous system that the Internet created.
The devices, the networks, the data streams — the entire kitchenware, as it were — are already in place, pre-heated and ready for this new recipe. This pre-existing infrastructure, coupled with a global population already accustomed to digital immersion, removes the natural friction that once slowed technological adoption.
The acceleration is not merely linear; it is geometric. This very fact places a Herculean burden upon governments, which must now comprehend and regulate a force whose ultimate societal and economic impacts are still unfathomable. To leave this force to its own devices, to allow a repeat of the laissez-faire approach of the 1990s, would be an act of profound historical negligence.
Successive global AI summits, commencing in the UK under then Prime Minister Rishi Sunak and progressing to France under President Macron, have served as vital, yet ultimately preliminary, diplomatic soundings. They have confirmed a universal recognition of the transformative power cascading from AI, but have thus far failed to forge the substantive, binding consensus required to steer this force.
The current landscape is a patchwork of fragmented and often contradictory approaches, a precarious situation that undermines global security and stalls innovation. The European Union, displaying characteristic regulatory ambition, has not waited for the dust to settle, enacting its comprehensive AI Act.
This is a significant step, but one taken largely in isolation. Across the Atlantic, the US maintains a stance of pronounced antagonism towards stringent regulation, favoring a sectoral, innovation-centric model that risks creating a dangerous regulatory vacuum.
Meanwhile, the technological reality accelerates, indifferent to these political machinations. The proliferation of foundational models, from ChatGPT and Claude to DeepSeek, has been succeeded by the deep integration of AI into the very operating systems of our lives through Apple Intelligence and Google Gemini. This is no longer a niche tool for developers; it is a pervasive layer being woven into the minutiae of human functions, from how we communicate and create to how we navigate our daily tasks.
The stage is, therefore, set for the third act in this consequential trilogy of conclaves, with the India AI-Impact Summit to be hosted in New Delhi on February 19-20, 2026.
The objective of the summit must be to lay down the foundational ground rules for the AI era. The objective is not, at this juncture, to draft a monolithic exhaustive legal code attempting to govern every conceivable application of AI.
Such an endeavor would be both futile and counterproductive, stifling the very innovation that promises so much progress. Instead, the aim must be to establish the first principles upon which all future national and international AI governance can be built.
It must be acknowledged that the AI-driven disruption is inevitable and, in many forms, desirable. The role of regulation is not to limit this growth, but to channel it, to ensure that the creative forces vastly outweigh the destructive ones, and that the benefits are distributed equitably across the globe.
The framework for these first principles must be built upon several core pillars. First, there must be a global commitment to safety, security and robustness, particularly for the most powerful, frontier AI models. This necessitates international standards for rigorous testing and evaluation, a shared understanding of critical risks, and protocols to prevent catastrophic misuse.
Second, the principles of transparency and explainability should be enshrined. Citizens and regulators alike must have insight into when and how AI systems are being used, especially in high-stakes domains like justice, finance and healthcare. Opaque algorithms making life-altering decisions are anathema to democratic accountability.
Third, a global consensus on data governance and privacy is paramount. The fuel for AI is data, and the unchecked harvesting and utilization of personal information without clear, harmonized rules represent a fundamental threat to individual autonomy.
Fourth, we must confront the profound ethical dimensions, establishing red lines against uses of AI that violate fundamental human rights, such as pervasive social scoring or lethal autonomous weapons systems operating without meaningful human control.
The term ‘to Google’ is on the cusp of becoming an archaic relic, a testament to the breathtaking velocity of this shift. This is merely the opening salvo, the software-based uprising. The true revolution, the confluence of AI with physical robotics and a fully realized Internet of Things, promises a radical reordering of our physical world within the coming decade itself.
(Manish Tewari is Lok Sabha MP and former I&B Minister)
