India Launches Its First Foundational AI Models at AI Impact Summit, Marking a Major Leap Toward Global Leadership
New Delhi: India unveiled its first foundational artificial intelligence (AI) models at the AI Impact Summit, signaling a significant stride toward establishing itself as a global leader in one of the world’s most transformative technologies.
The move is being viewed as a landmark moment in India’s technological journey, reflecting its ambition to build sovereign, large-scale AI systems capable of powering next-generation digital infrastructure.
What Are Foundational AI Models?
Foundational AI models form the backbone of modern artificial intelligence systems. These large-scale models are trained on vast amounts of data and can understand and generate text, images, audio, and other forms of content.
They serve as the core engine behind widely used technologies such as:
- Chatbots
- Voice assistants
- Image-generation applications
- Government AI platforms
In essence, foundational models act as the “roots” of AI ecosystems, enabling advanced applications to function efficiently and intelligently.
Artificial General Intelligence (AGI) is still years away
Speaking at the summit, Demis Hassabis, head of Google DeepMind, said that Artificial General Intelligence (AGI)—AI systems with human-like intelligence—remains five to eight years away.
He emphasized that despite rapid progress, AI systems still lack capabilities such as continuous learning and adaptable reasoning at human levels. “We are not there yet,” he noted, underscoring that significant technological breakthroughs are still required before AGI becomes a reality.
AI Mission 2.0 in the Works
Abhishek Singh, Additional Secretary at the Ministry of IT and CEO of the India AI Mission, announced plans for AI Mission 2.0. Under this initiative, a dedicated AI marketplace will be created where foundational models and applications developed in the first phase can be commercialized.
He added that feedback will be gathered from state governments and various departments to enable large-scale implementation of AI solutions across public services and governance systems.
India Moves Toward Sovereign AI Development
Bengaluru-based Sarvam AI introduced three foundational AI models with 3 billion, 30 billion, and 105 billion parameters.
According to the company, parameters represent the “brainpower” of an AI system—the higher the number of parameters, the more sophisticated and capable the model becomes.
In addition, BharatGen, a government-backed initiative associated with the Indian Institute of Technology Bombay, unveiled an early-stage 17-billion-parameter AI model.
Launched in October 2024, BharatGen aims to develop sovereign AI—indigenous AI systems owned and controlled by India.
The model supports multiple Indian languages, reinforcing inclusivity and national technological independence.
Union IT Secretary S. Krishnan highlighted that India has made strategic investments in building strong data systems to support AI innovation.
Google and Qualcomm Signal Strong Interest
Global technology leaders also expressed their commitment to India’s AI ambitions.
Sundar Pichai, CEO of Google, proposed deploying the company’s advanced AI models and products within Indian government agencies. He emphasized the potential of DeepMind’s AI technologies to improve public service delivery.
Meanwhile, Cristiano Amon, CEO of Qualcomm, stated that just as India progressed rapidly in the smartphone revolution, it is now poised to advance significantly in AI as well.
Qualcomm has earmarked ₹1,200 crore to support AI-related initiatives in the country.
India’s launch of foundational AI models, combined with government-backed sovereign AI initiatives and strong industry participation, marks a decisive step toward shaping the future of artificial intelligence on a global scale.
#IndiaAI #AIImpactSummit2026 #FoundationalAI #SovereignAI #SarvamAI #BharatGen #ArtificialIntelligence #DeepMind #DigitalIndia #TechInnovation
