news 
India Laud’s Chinese AI Lab DeepSeek, Plans to Host Its Models on Local Servers
India has expressed its admiration for the progress made by Chinese AI laboratory DeepSeek and announced its intention to host the lab’s large language models on domestic servers. The country’s IT minister, Ashwini Vaishnaw, praised DeepSeek during a conference on Thursday, stating that “You have seen what DeepSeek has done — $5.5 million and a very very powerful model.”
The decision to host DeepSeek’s models in India appears contingent upon the platform ensuring that all Indian users’ data is stored and processed domestically, aligning with India’s strict data localization requirements. This move marks a rare instance where China-based technology has been approved for use in India, given the country’s history of banning over 300 apps and services linked to China due to national security concerns.
DeepSeek’s eponymous model was recently brought to Microsoft’s cloud platform, sparking controversy surrounding its potential utilization of OpenAI APIs. However, the Indian government appears unphased by these concerns, instead focusing on the positive impact DeepSeek can have on the country’s burgeoning AI ecosystem.
The newly established computing facility will also support India’s broader AI initiatives, including 18 AI-driven applications aimed at addressing pressing issues such as agriculture, climate change, and learning disabilities. These projects are set to receive initial funding, with Vaishnaw emphasizing that “we will be keeping our models open and application-focused.”
India is actively working on developing its AI capabilities, with the IT minister estimating a projected investment of $30 billion in hyperscalers and data centers over the next two to three years. This endeavor will reportedly be overseen by a regulatory body utilizing a “hub-and-spoke model,” allowing multiple institutions to collaborate on safety frameworks.
The establishment of this facility is expected to have far-reaching implications for India’s tech industry, with Vaishnaw stating that 6-8 major developers are capable of creating AI models within six months to a year, and 4-6 within four to eight months.
Source: techcrunch.com
