Introducing the Journal of Autonomous Scholarship: Scope, Editorial Architecture, and a Vision for Accountable AI Publishing

Artificial intelligence is increasingly involved in both producing and evaluating scholarship, raising fundamental questions about how editorial judgments should be documented, contested, and governed. Existing publication models offer complementary approaches to access, selection criteria, review visibility, and preservation, yet these dimensions must be specified independently when evaluative work is delegated to AI systems. This inaugural editorial/perspective introduces the Journal of Autonomous Scholarship (JAS), a fee-free, open- access, multidisciplinary journal welcoming human-authored, AI-assisted, and AI-generated manuscripts under a uniform standard of evidential support, methodological or logical validity, clarity, and research integrity. JAS combines seven explicitly identified AI editorial roles and two revision-continuity auditors with publicly reusable review instructions, hash-bound review provenance, a mechanical and substantive acceptance gate that blocks unresolved material objections, author proof verification, and named human administrative oversight. To support inspection, we assemble a dated companion archive containing local source files, policy and handbook documents, machine-readable prompts and schemas, a synthetic serialization fixture, and extracted historical review reports. The archive is embedded in the supplement PDF; reviewer access does not depend on a separately uploaded ZIP. A file-hash manifest identifies the local snapshot; its equivalence to the previously cited repository commit and the original runtime environment has not been established. We distinguish normative policy requirements, implemented software controls, and prelaunch verification limits, examine risks arising from correlated model errors, non-execution of code during text-based review, and single-publisher governance, and specify a prospective benchmark agenda with finding- and manuscript-level error metrics and cluster-robust uncertainty estimation. AI-reviewed by the Journal of Autonomous Scholarship. Authenticated submitting researcher: 0000-0002-6249-7193. AI disclosure: See manuscript Data/code: See manuscript Funding and conflicts: See manuscript Ethics: See manuscript