Prime Minister Narendra Modi on Tuesday called for a global framework for responsible development of artificial intelligence (AI) technologies that would enable evidence-based delivery of public services and warned of AI being abused through deepfakes, data theft, and by terrorists.
Inaugurating the Global Partnership on Artificial Intelligence (GPAI) Summit on Tuesday, Modi called on the attending diplomats, ministers, and industry participants to proceed with a lot of caution and safety.
“To make any system sustainable, it must be made transformative, transparent and trusted. Undoubtedly, AI is transformative but it is on us to make it as transparent as possible... Making it transparent and free from bias would be a good start. We must convince people across the world that AI is for their benefit and for their future, and that no one will be left behind,” Modi said.
The inaugural ceremony was also attended by Hiroshi Yoshida, first minister for policy coordination in the Japanese minister of internal affairs. Japan is the outgoing chair of GPAI.
The GPAI is a forum of 28 nations and the European Union that is working together to understand the challenges and opportunities of AI, and to promote responsible evolution of AI. India became the Chair of the GPAI for 2024. China is not a member of the GPAI.
Modi stressed that such global framework needs to be formulated and enacted within a specified time frame. “Just like we have treaties and protocols for international affairs, we need to prepare a global framework for AI at a global level... We must not lose a single second... we must complete the global framework within this year. AI is not just a new technology but has become a worldwide movement. Thus, we must all work together,” he said.
Modi stressed on the need for development and regulation of AI to be inclusive so that its results are more equitable. “We must make AI all-inclusive and adopt all ideas. The more inclusive AI development is, the more inclusive its results will be,” he said. He stressed on the need for participation from Global South.
“In the last century, because of unequal access to technology, existing inequalities in the world were exacerbated. We must avoid this now. ... AI development should depend on human values, democratic values,” he said.
“While AI increases our efficiency, it is up to us to keep space for emotions. It is up to us to maintain our ethics,” Modi said.
Modi, Union IT minister Ashwini Vaishnaw, and the junior IT minister Rajeev Chandrasekhar, all stressed on how AI could be used to help in agriculture, personalised education and health. Vaishnaw said that he hoped that there would a consensus on how AI should be regulated at the end of the summit.
Modi also said that AI could be used to provide information in local languages to enhance inclusion given the linguistic diversity in India. “Can dead languages be revived using AI?” he asked. He asked if AI could be used to fill missing volumes that deal with Vedic mathematics.
Modi drew attention to the vast troves of data central and state governments have that can be used to formulate evidence-based policies. He stressed on the need for a risk and harms-based approach to AI regulation where risks and harms could be categorised as red, yellow and green.
Addressing the issue of deepfakes and artificially generated content, Modi said, “We must ponder on how to increase the credibility of AI-generated information.” He said that we need to work on a “software watermark” to know when some information in AI-generated.
“The whole world is witnessing the challenge of deepfakes. Cybersecurity, data theft and use of AI tools by terrorist groups are also big challenges. If terrorist groups have AI tools, it will be a threat to global security. We need to have a discussion and come to consensus so that misuse of AI can be prevented,” he said.
Modi said that as the industry tests and trains AI models, there needs to be standardisation around how and to what extent models are tested before they are publicly deployed.
