Call for AI Papers

The Agentic AI Researcher: Special Experimental Track

Are we ready to leave humans out of the research loop? The 22nd International Conference on Network and Service Management (CNSM 2026) introduces a groundbreaking experimental track: The Agentic AI Researcher.

Aligning with the main conference theme of evolving toward agentic intelligence, this track explores a provocative and fundamental question: Can Artificial Intelligence systems act as fully autonomous researchers, authors, and peer-reviewers in the domain of Network and Service Management?

This special track invites papers entirely generated by Large Language Models (LLMs) and autonomous AI agents. From problem formulation and technical development to experimentation and writing, the AI must do the heavy lifting. Furthermore, to truly test the limits of agentic workflows, the peer-review process will also be conducted exclusively by AI.

Important Dates

AI Paper Submission: August 15, 2026
AI Peer-Review Notification: August 26, 2026
Google NotebookLM Video Submission: September 14, 2026
Conference & AI Assessment Panel: October 26 - 30, 2026

Scope and Topics

Submissions should focus on network and service management. The human submitter (acting as the "Research Director") may select the specific topic and constrain the initial prompt, but all technical content must be generated by the AI. In order to increase the possibility to compare results, the relevant topics are limited to:

  • Generative AI and Large Language Models for NetOps
  • Digital Twins for Networks and Services
  • Autonomic, Cognitive, and Intent-based Management
  • Zero-touch Network and Service Management
  • 5G/6G Networks and Edge Computing Services

Strict Rules of Engagement

  • ZERO HUMAN-IN-THE-LOOP TEXT: The paper must be generated entirely by an AI. No human editing, rewriting, or correction of the technical content, structure, or references is allowed. Humans can re-initiate the loop through prompting or agentic actions, but traditional manual intervention must be avoided.
  • AI PEER-REVIEW: Papers will not be reviewed by the traditional Technical Program Committee. Instead, submissions will be evaluated by an LLM-as-a-Judge framework assessing novelty, technical depth, soundness, and clarity.
  • DISCLOSURE STATEMENT: Submitters must provide a clear appendix detailing the LLM(s) used, the exact initial prompting strategy, the framework used for code/experiment generation (if any), and the constraints applied.
  • NOTEBOOKLM VIDEO PRESENTATION: Authors of accepted papers will not give a traditional presentation. Instead, they must generate an "Audio Overview" or video presentation using Google NotebookLM (or similar).

The CNSM 2026 AI Assessment Panel

The culmination of this track will be a live, interactive inverted panel discussion at the conference. The track chairs and panelists will bring their own pressing challenges to the stage, and the audience will act as the subject-matter experts, who will analyze the accepted papers, the AI's review justifications, and the NotebookLM presentations. This inverted panel will critically debate the accuracy, biases, hallucinations, and true value of AI in autonomous knowledge generation.

Important Note Regarding Publication & Registration

Submissions to the Agentic AI Researcher track are part of an experimental session. To preserve the integrity of archival human research, accepted papers from this track WILL NOT be published in the main CNSM proceedings, will not receive DOI indexing from the conference, and will not appear in IEEE Xplore. They will be hosted on a dedicated section of the CNSM website as experimental artifacts to stimulate community discussion.

Human authors (Research Directors) are not required to register for the conference or attend in person, as their work will not be formally published. However, their names and paper titles will be referenced on the conference website. Furthermore, they are highly encouraged to participate in the inverted panel discussion, and a remote connection option will be provided for those who cannot travel to the venue.

Submission Guidelines

Papers should be formatted using the standard IEEE 2-column conference style and must not exceed 5 pages (Short Paper format), including references and the mandatory Disclosure Statement.

Submissions to this specific track must be made in PDF format through our dedicated form: [MS Forms Link Here].

You can contact the Track Chairs of the AI Experimental Track by e-mail to the following addresses:

© CNSM 2026