India AI Impact Summit 2026: India Must Build Sovereign AI Path, Not Copy EU Or US, Say Experts
Experts backed light-touch governance, domestic capability-building, and trusted infrastructure to balance innovation, security and growth.
Published : February 16, 2026 at 4:35 PM IST
By Gautam Debroy
New Delhi: The India-AI Impact Summit 2026 is underway in Delhi, bringing a cohort of tech and world leaders to the national capital. Considered one of the world's largest and most exhaustive gatherings on AI so far, the Summit places India at a crucial juncture in understanding and charting a course on how to use and leverage this disruptive technology.
Speaking to ETV Bharat, AI and technology experts said India's AI moment must be rooted in strategic autonomy. They stressed that India should maintain innovation speed without copying the European Union's regulation-heavy model or the United States' market-driven approach.
Golok Kumar Simli, former Chief Technology Officer (CTO) for the Passport Seva Programme under the Ministry of External Affairs (MEA), told ETV Bharat that India can truly chart a third path.
"The path is an adoption-first, sovereignty-backed framework built on four strategic layers – Chips, Cloud, AI, and Knowledge that I fondly term as a 'Sovereignty Pyramid'," he said.
Operationalising this 'Sovereignty Pyramid' would, he said, enable India to become not just an AI adopter, but also a global standard-setter.
He suggested focusing on securing the compute backbone, which includes accelerating semiconductor fabrication, advanced packaging, and AI accelerator ecosystems.
"We may link incentives (PLI, ISM) to domestic R&D centres, IP co-creation, and skills transfer. We need to build strategic reserves and diversified supply partnerships. Without chip resilience, cloud and model ambitions remain dependent as AI sovereignty begins with computer control," Simli said.
Simli, who is presently engaged with BLS International as its president, technology and innovation, said trusted digital infrastructure is essential and called for a nationally certified Trusted Cloud & Compute Framework and encryption key localisation for critical sectors.
He added that India needs to establish a framework for audited cross-border data flows under the DPDP Act and create sovereign AI training environments for public-sector use. He said balanced data protection is vital to building citizen trust while maintaining interoperability for global trade and collaboration.
AI & Knowledge Sovereignty
Emphasising the need for what he described as innovation backed by safeguards, Simli said, "Innovation with guardrails – crucial is, how do we launch publicly funded AI model sandboxes with subsidised compute credits, implement risk-graded AI regulation with strict oversight for high-risk use cases (justice, health, surveillance) and light-touch norms for low-risk innovation.”
He suggested that India must ensure that transparency, auditability, and impact assessments for critical AI systems are a mandate and not an afterthought.
"We may combine EU-style clarity for high-risk AI with US-style innovation incentives without overregulating startups."
Highlighting the importance of data, skills and standards, Simli said India must develop privacy-preserving national data commons (multilingual, agriculture, health, governance). Scaling AI skilling missions across states is also an essential element, he said.
"We have to promote open standards and India-led AI governance templates for the Global South. To me, true sovereignty lies in intellectual capital, domain datasets, and standard-setting influence."
According to Simli, the strategic outcome must align industrial policy, digital public infrastructure and responsible AI governance.
"This would help us safeguard national security and citizen rights, accelerate domestic AI innovation, reduce strategic technology dependence and shape global AI governance norms. In essence, India’s AI future should not be framed as regulation versus innovation. It must be designed as sovereignty enabling innovation that touches the common mass too with equal ease and accessibility,” he said.
India’s Evolving AI Governance Framework
Speaking to ETV Bharat, Subimal Bhattacharjee, columnist and author of the book The Digital Decades: Thirty Years of the Internet in India, and former country head at General Dynamics, said India released comprehensive AI Governance Guidelines in November 2025, marking a major milestone in balancing innovation with safeguards.
He said the guidelines are distinctive in several ways, as India has adopted a 'light-touch' risk-based approach that prioritises innovation over restraint, using voluntary measures, digital public infrastructure (DPI), and a techno-legal strategy.
The framework, Bhattacharjee said, encourages compliance through voluntary measures supported by techno-legal solutions, with additional obligations for risk mitigation applying in specific contexts.
He said that India's AI Governance Framework is based on seven guiding principles: trust, human-centrism, responsible innovation, fairness and equity, accountability, understandability by design, and safety, resilience, and sustainability.
"India focuses on voluntary commitments, self-certifications, transparency reports, and third-party audits before imposing strict responsibilities, with incentives like sandbox access, reputational badges, and targeted support. This reflects India’s developmental priorities and its goal to leverage AI for inclusive growth while building domestic AI capabilities," he said.
European Union's Comprehensive Regulatory Framework
According to Bhattacharje, the EU has taken a fundamentally different approach with the AI Act.
“The AI Act entered into force on August 1, 2024, and will be fully applicable by August 2, 2026, with maximum penalties up to €35 million or 7 per cent of worldwide annual turnover, whichever is higher. The EU establishes four risk levels: unacceptable risk (banned practices like social scoring), high risk (strict requirements for systems in areas like employment, law enforcement, and essential services), limited risk (transparency requirements), and minimal risk (no obligations),” he said.
United States' Fragmented And Evolving Approach
Bhattacharjee said the US presents the most complex picture, with recent dramatic shifts in federal policy and ongoing tension between federal and state regulation.
He noted that in December 2025, President Donald Trump issued an executive order titled “Ensuring a National Policy Framework for Artificial Intelligence,” revoking the earlier directive and seeking a minimally burdensome national standard instead of multiple state regimes.
“More than 1,000 AI-related bills have been introduced across nearly every state in 2024-2025, with states like California, Colorado, New York, and Illinois adopting or proposing comprehensive AI and algorithmic accountability laws. However, the executive order directs the Attorney General to establish an AI Litigation Task Force to challenge state AI laws on grounds including unconstitutional burdens on interstate commerce and federal preemption,” he said.
According to Bhattacharjee, the order also asks the Secretary of Commerce to evaluate “onerous” state laws and to make those states ineligible for certain federal funding.
“The order declares that ‘United States AI companies must be free to innovate without cumbersome regulation’ and positions this as necessary for maintaining US leadership in an AI race with adversaries,” he said.
Bhattacharjee said India strikes a middle path while building institutions. “EU prioritises safety and fundamental rights protection with mandatory compliance and significant penalties, and on the other hand, the US currently emphasises minimal regulation and industry self-governance, actively opposing state-level restrictions,” he said.
According to him, India reflects its position as a developing economy.
Also Read:
