India is accelerating efforts to develop its own artificial intelligence (AI) ecosystem, with the government selecting 18 projects to build foundational AI models, technology minister Ashwini Vaishnaw said on Thursday.

Six major developers have been identified to build these models, which are expected to be ready within eight to 10 months. As part of the IndiaAI Mission, the government has secured 18,693 graphics processing units (GPUs) to power AI research and development, exceeding its initial target of 10,000 GPUs.

The computing infrastructure, provided by companies including Jio Platforms, Tata Communications, and Yotta Data Services, will be available to start-ups and researchers at a highly subsidised cost, Vaishnaw said.

“We will build a world-class foundational model that can compete with the best in the world,” Vaishnaw said, adding that six major developers would be able to launch AI models within eight to 10 months. Some could be ready in as little as four months.

Foundation models, the core of generative AI applications, are large, pre-trained neural networks that power AI-driven solutions. India’s models will be developed using datasets “for our country, of our country, and for our citizens,” ensuring they address local challenges and remove algorithmic biases, the minister said.

Altman visit

Microsoft-backed OpenAI’s chief Sam Altman is planning to visit India on February 5, three sources with direct knowledge of the matter said, in what could be his first visit in two years at a time the company faces legal challenges in the country.

One of the sources said a meeting with government officials was also on the cards.