The annual Global Partnership for Artificial Intelligence (GPAI) Summit kicked off in New Delhi’s Bharat Mandapam on December 12, with discussions on AI issues like safety and development challenges.
“Trust in AI will grow only when related ethical, economic and social aspects are addressed,” Prime Minister Narendra Modi said while inaugurating the three-day event. 29 countries, from North and South America, Europe and Asia are part of GPAI. China is not a member; the previous summit was held in Osaka, Japan, and India will be the lead chair of the grouping in 2024.
Other discussions throughout the day featured AI policymakers and industry officials presenting their views and expertise. Union Minister for Electronics and Information Technology Ashwini Vaishnaw had said on December 11 that issues related to farming and healthcare would be discussed thoroughly at the summit.
“In India, farmers are the least data-integrated,” Vijay Shekhar Sharma, Paytm founder and CEO said, while adding that “AI will bring the nuances of the land, the ecology and the kind of produce that you can build, and the demand-supply gaps we can fulfil”. Ecological issues would arise from a less informed ‘lowest common denominator’ approach to farming, Mr. Sharma said.
Other issues discussed included data sharing and cross-border data flows, issues actively being deliberated by the Union Government as it tries to rebalance private firms’ power when it comes to ownership of data. “Cross-border data flows is one area where we certainly need to look into it, because when we talk about bringing the equitable compute, then how do we bring the data, and who owns the data are very important aspects,” Kavita Bhatia, a group coordinator at the IT Ministry said at a panel discussion, adding that the issue would be deliberated at the data governance working group of GPAI.
Mr. Modi also raised other pertinent issues surrounding the AI debate. “Can we establish an institutional mechanism that ensures resilient employment? Can we bring standardised global AI education curriculum? Can we set standards to prepare people for an AI-driven future?” he asked. He also invoked watermarking of AI products, an apparent reference to suggestions that imagery generated by AI applications be distinguished as such. AI-generated images have been used in illicit ways such as creating ‘deepfake’ clips of celebrities, as well as by firms seeking to save costs by creating photorealistic images for advertisements.
Mr. Modi also pointed to the dangers of terrorists getting access to AI tools, as well as the potential for increased cybersecurity and data theft incidents. He called for visibility into the algorithms that go into creating AI models. “There is no doubt that AI is transformative but it is up to us to make it more and more transparent,” he said.