SYSTEM NOTICE

Auto translation by AI. Be sure, accuracy, nuances and authorial intent may not be fully reflected.
見出し画像

Why is AI Agent Implementation Lagging Despite Their Increased Intelligence? — From the Opening Remarks of the NRAC AI Agentic Summit 2025

On November 27, 2025, the "NRAC AI Agentic Summit 2025" opened at Otemachi Mitsui Hall. In his opening remarks, Koichi Hasegawa, Chairman of the Next-Generation RPA/AI Consortium (NRAC), raised essential issues regarding the utilization of AI agents in Japan.


Opening Remarks by NRAC Chairman Hasegawa

The Reality of AI Agent Implementation: POCs are Progressing, but Implementation is at 10%

Chairman Hasegawa began by presenting shocking data.

"Agents remain at the POC stage, and cases where they are actually utilized in business are only about 10%."

With the evolution of generative AI, the "brains" of AI agents are becoming smarter every day. However, implementation in actual business operations is significantly lagging—this gap is the biggest challenge facing Japanese companies.

Why is implementation not progressing? Three hypotheses

Chairman Hasegawa presented three possibilities behind this issue.

1. Weakness in the action (execution) part

Although the "brains" of AI agents are developing, the execution functions that act as the "hands and feet" to actually carry out tasks may not be keeping up.

There is a possibility that the execution layer, centered on Robotic Process Automation (RPA), is not yet sufficiently mature.

2. Lack of process design

Perhaps the entire business process has not been redesigned from an agent's perspective.

Trying to apply existing business workflows directly to AI agents does not work well. It is necessary to reconstruct processes that leverage the characteristics of agents.

3. Challenges in orchestration

After an action occurs, the AI agent makes a judgment, and a human provides final approval—perhaps such complex workflows are not being properly orchestrated (integrated management and coordination).

There is a need for a mechanism that clarifies the division of roles between AI, humans, and RPA, and links them organically.

What we aim for at today's summit

Chairman Hasegawa described the purpose of today's summit as follows.

Together with experts like Professor Sudo and Professor Nishimura, as well as practitioners from the industry, I would like to think about what we, as NRAC members and Japanese society, should do to effectively utilize Agentic AI next year.

The Path to Agentic Automation

Today's summit will discuss not just AI as a simple efficiency tool, but 'Agentic Automation'—a new collaborative model that integrates the 'thinking ability' of AI agents, the 'execution power' of RPA, and the 'judgment' of humans.

What is necessary for Japan to lead the world in AI utilization?

A day to explore that answer has now begun.

Event Overview
NRAC AI Agentic Summit 2025
Date and Time: Thursday, November 27, 2025, 12:30-19:00
Venue: Otemachi Mitsui Hall
Organizer: Next Generation RPA/AI Consortium (NRAC)

We are distributing the 'Next Generation RPA/AI White Paper 2025' free of charge to all attendees today.

#NRAC #AIAgent #AgenticAutomation #GenerativeAI #RPA #DX #AISummit

いいなと思ったら応援しよう!