[3L2-GS-5f] Agents

11:00 AM - 11:15 AM JST(2:00 AM - 2:15 AM UTC)

[3L2-GS-5f-01] Problems in Social Use of Agentic AI

〇HIROSHI NAKAGAWA¹ (1. RIKEN)

Keywords:

AI agent,Agentic AI,Social Application,trust,prompt

Research into Agentic AI, or AI agents, is progressing rapidly as an important application of generative AI. Agentic AI goes beyond providing information like generative AI; it performs a wide range of practical tasks according to the user's requests, such as purchasing, reservations, and contracts for personal use, and planning and decision-making, creating and implementing development plans, and external negotiations for corporate use. These tasks are expected to greatly contribute to expanding the capabilities of individuals and streamlining and streamlining corporate activities. However, since they involve practical work and internal corporate decision-making, many problems can arise, and in order to trust and utilize Agentic AI, it is necessary to identify these problems and develop solutions. In this presentation, we will report the results of a survey and analysis of the many papers recently published on this issue, and organize the key points.