INDIAai may submit AI rules framework in two weeks: Officials

The seven working groups constituted under India’s National programme for Artificial Intelligence are likely to submit recommendations for a comprehensive framework governing various facets of AI in the next two weeks, top officials told ET.

The working groups, set up under the aegis of the central programme that has now been christened as INDIAai, have been tasked with creating a framework for data governance framework, setting up an India Data Management Office as well as looking into the regulatory aspects of AI.

“The working groups have been constituted with representatives from industry and academia to fine tune the policy so that it’s made more realistic,” said one of the officials. Once the committees submit their report to the Ministry of Electronics and Information Technology (Meity) “it will be combined into a single document and reviewed before being implemented,” the person added.

The groups are also required to evaluate methods for capacity building, skilling and promoting startups in AI. They will also be required to guide moonshot or innovative projects in AI, the setting up of data labs among others.

In a recent interview with ET, minister of state for electronics and IT Rajeev Chandrasekhar said that “India is certainly going to be one of the largest data sets in the world. What makes India so attractive for AI researchers is the diversity of it. You have so much diversity and the quality of the data sets for any large language model or any AI learning model will be so exciting.”

While presenting the Union Budget 2023-24, finance minister Nirmala Sitharaman had announced setting up three Centres of Excellence for Artificial Intelligence in top educational institutions with an aim to help research and development of public-facing AI-based applications in agriculture, finance, health and other areas. Some of the working groups are also looking at finalising the contours of the centres of excellence. “India has the opportunity to make big strides in AI, with the potential to catalyse transformative changes across the country. Especially as the revolution in Generative AI unfolds, the government will need to protect and prepare the nation for what could be a very exciting opportunity replete with significant risks,” Srikanth Velamakanni, Group CEO Fractal Analytics told ET.

An important pillar of the India AI initiative is the India dataset programme, which aims to create a common database held across various government arms for the use of Indian startups and researchers. The programme will also give preferential access to the datasets to Indian startups.

“What we are saying is that having Centers of Excellence is one element, startups are one element, but the actual raw material for India AI and all the other learning models will be the India data sets program that will keep growing, we will we will accumulate all of what the government has and then as the government digitizes, you will get all kinds of other diverse data sets,” Chandrasekhar had added.

“One of the groups that has been formed is designing what the India datasets program would look like, what the platform would look like and what kind of datasets will be combined from the government with the private sector data,” a senior government official said.

Under the INDIAai plan, the IT ministry plans to establish Centres of Excellence, give exclusive access to Indian startups and Indian researchers to the datasets being collected by the government, and combine all of these to develop learning models to train AI for various public-centre projects, the official said.

“The current pace of innovation in AI offers India an unique opportunity to leapfrog AI adoption for inclusion and augmentation. The IndiaAI initiative by the government will help build the foundation through access to open data, AI skilling, Responsible AI practices and projects like Bhashini which will enable the benefits of AI to be available in multiple languages,” said Sangeeta Gupta, chief strategy officer of Nasscom, who is part of one of the working groups.

ET had on May 13 reported that the Centre plans to establish “some principles” which will act as “guardrails” in the AI sector which will also help regulate fast-growing language learning models such as Microsoft’s Open AI and Google’s Bard as well as their use by other companies.

Stay on top of technology and startup news that matters. Subscribe to our daily newsletter for the latest and must-read tech news, delivered straight to your inbox.