Deliberating AI in India: Insights from a National Deliberation
Zachary Serban (Stanford) - United States
Keywords: india, deliberative poll, ai, technology
Abstract
This paper examines the perspectives of Indian participants on artificial intelligence (AI) through a national Deliberative Poll. A baseline survey conducted with 1,068 respondents identified key trends and opinions on AI, which were further explored through a Community Forum. Of the 570 participants in the treatment and control groups, 271 completed the post-survey after deliberation. Before deliberation, participants expressed high excitement about using AI for information gathering but were least excited about social interaction. Post-deliberation, participants showed a notable increase in support for proposals that granted users control over AI’s memory, particularly the ability to select and update data settings, which saw the highest increase in approval (17.9%). Support for AI’s proactive role in nudging users also grew significantly, while views on the autonomy of AI agents decreased when transparency was not ensured. While Indian participants largely supported AI’s role in personalizing experiences, they were more cautious about AI's role in human-like interactions and relationships. After deliberations, participants demonstrated an improved understanding of AI, as evidenced by a notable increase in correct responses to factual questions about AI. Demographic analysis showed that age was a significant factor in differentiating opinions, while gender did not emerge as a differentiating factor. The findings highlight the evolving and nuanced views on AI in India, underscoring the importance of transparency, user autonomy, and education in shaping public opinion.