Accountability

Human authorship, transparent AI assistance.

Named authors are natural persons who review the work, accept responsibility for its claims, and remain accountable for corrections. AI systems are tools—not authors, committers, or co-authors.

Authorship boundary

Codex, Claude, Grok, and other generative systems receive no authorship credit.

Git metadata

AI systems must not be listed as a Git author, committer, or Co-authored-by identity. Repository checks inspect new commit ranges for prohibited machine attribution.

Research outputs

Scholarly authorship belongs only to people who meet the target venue's criteria and accept responsibility for the submitted work.

Transparent disclosure

Material AI assistance is recorded in provenance, methods, acknowledgements, or limitations where it helps a reviewer understand how the work was produced.

Human responsibility

A policy against machine co-authorship does not permit unreviewed output or false human attribution. The named human remains responsible for evidence, claims, and corrections.

Repository control

A visible policy plus an automated guard.

The reusable check rejects new commits when author, committer, or co-author metadata identifies Codex, Claude, Grok, OpenAI, Anthropic, or xAI. It is intended to be required on protected default branches. GitHub plan limitations and legacy history are reported rather than hidden.

Authorship is accountability, not software attribution.Return to About Request clarification