IndiaAI Mission: MeitY ramps up efforts to ensure availability of AI-usable data across govt

The India Datasets Platform will be established as part of the Rs 10,732 crore IndiaAI Mission

As part of the IndiaAI mission, the Ministry of Electronics and Information Technology is currently focusing on building capacities across the government to ensure the availability of artificial intelligence-usable datasets on the proposed India Datasets Platform, said Abhishek Singh, additional secretary in the ministry.

Under IndiaAI Mission, the government aims to streamline access to non-personal datasets to entrepreneurs and startups. The Rs 10,732-crore initiative also aims to provide access to compute infrastructure to startups, researchers, create AI use cases and more.

Singh, who is also the CEO of the IndiaAI Mission said, "We are building capacities within each department (ministries), whether it's health, agriculture, education, to ensure that they are able to work out the mechanism that use the policies which are there, and put data on the India Datasets Platform to make it usable."

He was speaking at at the Telangana Govt's Global AI Summit on September 5.

Also read: Decoding IndiaAI: Inside India's Rs 10,372-crore plan to boost AI infrastructure

Singh said that under the India Datasets Platform, the government will set up a India data management office, which will ensure that all necessary policy and tools are in place to ensure that data is kept in ready a form that is usable by AI.

"Then we also have data management units in every government department. Like for example, today health has a lot of data, but they don't have any capacity within the health ministry for deciding which data I should put in the open domain, which data I should put in restricted domain and which data I should not put at all," he explained.

Despite the Open Data Portal of the Government of India boasting over 5,00,000 datasets, Singh highlighted that only a small fraction of these are actually usable for AI.

"We might have lots of datasets.. But how many of it is actually usable for the AI ecosystem? Very few. There are challenges there with regard to not having confirming to data standards, metadata standards, privacy preservation tools, not having anonymisation and so on. So there are multiple challenges there," he said.

Last month, the IT ministry, invited applications from to procure 10,000 GPUs to provide access to artificial intelligence (AI) services over cloud to academia, MSMEs, startups, governments, and other public sector agencies, under the IndiaAI Mission.

A week later, in an jam-packed pre-bid meeting for procuring the GPUs , industry stakeholders sought a number of clarifications from the IT ministry, including requesting that the requirements in the tender be modified so that it encourages smaller players to participate in the empanelment process, three sources aware of the matter said..