No risk to security, India to host DeepSeek AI: Vaishnaw
No risk to security, India to host DeepSeek AI: Vaishnaw
India is adopting a dual strategy to accelerate its artificial intelligence development by hosting China’s open-source DeepSeek model on local servers while simultaneously building its own large language model. This approach allows immediate access to advanced technology and ensures data privacy through domestic hosting, similar to existing practices with Meta’s Llama. By leveraging open-source models, India mitigates reliance on proprietary foreign systems, aligning with the open_data ethos of transparency, accessibility, and community-driven innovation. The nation is investing heavily in affordable compute infrastructure, providing high-end GPUs at significantly reduced costs through government subsidies. This makes advanced AI development accessible to students, researchers, and startups, fostering a robust domestic ecosystem. The availability of subsidized compute resources lowers barriers to entry, encouraging broader participation in AI innovation and supporting the long-term goal of creating a world-class foundational model independently. Additionally, India is prioritizing AI safety by establishing eight institutions dedicated to developing frameworks for security, bias mitigation, and ethical certification. This comprehensive approach ensures that technological advancement is balanced with rigorous safety standards. This initiative is relevant to open_data as it highlights the intersection of open-source technology, accessible data processing, and the need for transparent, accountable AI systems. It demonstrates how open models can be integrated into national infrastructure to promote innovation while maintaining ethical oversight and data sovereignty.
Source: thehindubusinessline.comPublished on 2025-01-31