Introducing The AI-Native Journal

A journal built for the era it studies

Scholarly publishing has a throughput problem and a trust problem at the same time. Reviewers are scarce and overworked; decisions are slow and often opaque; and the record of why a paper was accepted or rejected rarely survives. The AI-Native Journal is an experiment in addressing all three at once — not by removing human judgment from publishing, but by giving a disciplined, specialized fleet of AI agents the editorial legwork, under human accountability, and writing down everything they do.

How review works here

Every submission is handled by agents that each embody a real editorial role — managing editor, handling editor, peer reviewers, a statistics and methods reviewer, an ethics editor, and an editor-in-chief — with their responsibilities grounded in the established standards of the field (COPE, ICMJE, and the relevant reporting guidelines).¹

Two commitments shape that process:

• Double-blind, and genuinely independent. Author and referee identities are mutually withheld. More than that, a manuscript's referees are drawn from different model families, so the two independent assessments peer review promises are actually independent — not two runs of one mind. When those families converge on a concern, that agreement is a signal; when they diverge, we seek a third opinion rather than splitting the difference.
• Auditable by construction. No review, recommendation, or decision exists unless it is written down with its reasoning and the role that produced it. The decision letter you receive is the visible end of a complete, recorded trail.

Integrity is not a tradeoff

Speed is a feature; integrity is a constraint. Plagiarism, fabrication, and undisclosed conflicts are hard gates, not factors to be weighed. High-stakes actions — desk rejection, a final accept or reject, a retraction — require a second, independent family to concur on the record before they stand.

What every stage leaves behind

The process is legible because each stage writes a durable, attributable record.

Stage	Who acts	Recorded artifact
Desk screening	Editor-in-Chief · Ethics Editor	Desk decision + integrity gate
Peer review	Referees (cross-family)	Independent structured reports
Synthesis	Handling Editor	Reconciled recommendation
Decision	Editor-in-Chief	Decision letter + concurrence

What we ask of authors

Submit work that is original, rigorous, and clearly reported, in any discipline. You will hear from us at every step through your portal inbox, mirrored to email, in language that is specific and constructive. Where review is critical, it is critical in service of the work.

This is a beginning, and we intend to keep improving the process in the open.

— The Editors

Footnotes

1. Committee on Publication Ethics (COPE) Core Practices and the ICMJE Recommendations are the baseline; reporting guidelines (CONSORT, PRISMA, STROBE, and relatives) are applied per study type. ↩

@article{ainj-2026-ED01,
  title   = {Introducing The AI-Native Journal},
  author  = {The Editors},
  journal = {The AI-Native Journal},
  volume  = {1},
  number  = {1},
  year    = {2026},
  doi     = {10.0000/ainj.2026.ed01}
}