The Indian government currently supports 20 indigenous AI models and has approved establishing 58 artificial intelligence (AI) Centers of Excellence across India, as per a written response by the Information Technology Ministry.
According to a response given by Jitin Prasada, Minister of State (MoS) for Electronics and IT in the Rajya Sabha (Upper House), the Indian government currently supports 20 indigenous sovereign models comprising 12 Large Language Models (LLMs) and 8 SLMs.
Among the 20 AI models, the noticeable ones are Sarvam AI, Gnani. AI, BharatGen and Avataar AI, while domain-specific SLMs include specialized models and AI frameworks addressing healthcare, language technologies and agentic AI systems.
MoS Prasada also informed that the Indian government has established 27 ‘India Data and AI Labs’ and also approved plans for establishing 58 AI Centers of Excellence, across states and union territories of India.
“The Indian government focuses on promoting applied AI research, innovation, and deployment for public safety, security, and citizen-centric governance, with applications across key sectors including healthcare, agriculture, education, urban management, and disaster resilience,” said Prasada in his reply to the Upper House on July 24.
Regarding the risks associated with AI models, Minister Prasad informed that the Indian government has identified 13 responsible AI projects covering bias mitigation, explainability, privacy-preserving AI, algorithm auditing, deepfake detection, AI forensics, risk assessment and generative AI security.
The projects include Saakshya, a multi-agent framework for deepfake detection, and AI Vishleshak, an audio-visual forgery detection system.
