Generative AI Policy

Authors may use generative AI tools with care, but they must disclose that use, and they stay fully responsible for every part of their manuscript. AI tools cannot be authors. This policy follows the COPE position on authorship and AI tools.

• No AI authorship: AI tools cannot meet the authorship criteria in Authorship and Contributorship. They cannot take responsibility for the work, approve it, or declare competing interests. They must not be listed as authors, or cited as authors.
• Permitted uses with disclosure: improving the language and readability of the authors' own text; helping to write or debug code; generating ideas for structure. Whatever the use, the authors must check every output for accuracy, bias, and plagiarism.
• Prohibited uses: generating or altering research data, results, or images; creating references that the authors have not read and checked; producing a substantial part of the scientific content (analysis, interpretation, or conclusions) without real intellectual work by the authors.
• Exception for research methods: if AI is part of the research itself, for example a machine-learning model used in the study, it is described in the Methods. This includes the model name, version, settings, and how it was validated.
• Disclosure: a section headed "Use of Generative AI", placed before the references, names each tool and version, what it was used for, and which parts of the manuscript it affected. For example: "During the preparation of this work the authors used [tool, version] to improve the readability of the Introduction. The authors reviewed and edited the output and take full responsibility for the content." Ordinary spelling and grammar checkers do not need to be disclosed.
• SSG's own tools: authors must disclose the use of generative AI tools in preparing their manuscript, including SSG's own StepSphere, as described in this policy.
• Undisclosed use: if undisclosed AI use is found, or AI-generated content that is false (for example, references that do not exist), it is handled as misconduct under Publication Ethics, section 2.

Manuscripts under review are confidential. Reviewers must not upload a manuscript, or any part of it, or their own review, into a generative AI tool. Doing so may break confidentiality, data protection, and the authors' rights. This includes SSG's StepSphere. StepSphere's terms also prohibit such use. Reviewers may use AI only to improve the language of their own report, and only without including any content from the manuscript. The scientific judgement in a review must be the reviewer's own.

Editors must not upload manuscripts, reviews, or decision letters into generative AI tools. This includes SSG's StepSphere. StepSphere's terms also prohibit such use. SSG may use secure tools that the publisher has approved, for technical checks such as similarity screening, reference checking, or image screening. Every decision is still made by a human editor.

SSG's StepSphere workspace is a commercial service. It is fully separate from the editorial process. Using it does not affect how a manuscript is handled or decided. The editorial office does not use StepSphere to assess submissions. See the StepSphere Policy for how the service works and how it is kept separate from SSG journals.