Artificial Intelligence & Human Authorship Verification Policy

Effective Date: 11 September 2026
Policy Owner: Tundra Swan Press
Applicability: This policy applies to authors, editors, contributors, translators, illustrators, and other creators submitting material to Tundra Swan Press (TSP).

1. Purpose

Tundra Swan Press is committed to publishing work that reflects genuine human creativity, authorship, judgment, and responsibility.

Artificial Intelligence (AI) tools can assist with research, brainstorming, editing, accessibility, translation, formatting, and various other legitimate elements of creative and publishing tasks. At the same time, undisclosed or substantial AI-generated content can raise questions concerning authorship, originality, copyright, accuracy, artistic integrity, and the reliability of representations made to TSP.

This policy establishes Tundra Swan Press’s standards for:

  • Human authorship and creative responsibility.
  • The disclosure and appropriate use of generative AI.
  • Verification of authorship when questions arise.
  • Editorial review of suspected undisclosed AI-generated material.
  • Fair, proportionate responses to potential violations.

This policy is intended to protect both the integrity of Tundra Swan Press’s publications and authors who use AI tools responsibly and transparently.

2. Core Principle: Human Authorship

Tundra Swan Press publishes works for which a human author or authors retain creative responsibility. Authors are responsible for the work they submit, including its:

  • Content and meaning.
  • Originality.
  • Representations concerning authorship; and
  • Final form submitted for publication.
  • Factual accuracy.
  • Citations and attributions when applicable.
  • Legal and ethical compliance.

The use of an AI tool does not, by itself, determine whether a work is considered human-authored. The important question is the nature, extent, and creative significance of an AI’s contribution and the author’s subsequent human contribution and control.

An author may use AI as a tool without surrendering authorship. However, a work that is in-part or substantially generated by AI and merely reviewed, lightly edited, or presented as the author’s original human creation will not satisfy Tundra Swan Press’s authorship standards.

3. Definition of AI

For purposes of this policy, AI includes generative artificial intelligence and related automated systems capable of producing, transforming, or substantially altering text, images, audio, video, code, translations, or other creative material. Examples include, but are not limited to:

  • Large language models and AI writing assistants.
  • Generative image systems.
  • AI voice and audio-generation systems.
  • AI video-generation systems.
  • Automated translation or rewriting systems.
  • AI-powered editing or content-generation platforms, and
  • Other systems that produce or significantly transform creative content in response to user instructions.

Ordinary software functions such as spell-checking, grammar correction, formatting, search, or conventional word-processing features are not ordinarily considered generative AI for purposes of disclosure unless they materially generate or transform the author’s creative expression.

4. Permitted AI Use

Tundra Swan Press permits responsible use of AI tools, provided that the resulting work complies with this policy and the author remains implicitly responsible for the creative work. Permitted uses may include:

4.1 Brainstorming and Ideation: Authors may use AI to generate ideas, questions, outlines, prompts, possible titles, or other preliminary material, provided that the published work represents the author’s own creative contribution and no generative AI content was cut and pasted into the work.

4.2 Research Assistance: AI may be used as a research aid or to help identify subjects for further investigation. Authors remain responsible for independently verifying facts, quotations, citations, sources, names, dates, statistics, and other information.

AI-generated factual claims should not be treated as authoritative merely because an AI system produced them.

4.3 Editing and Revision Assistance: Authors may use AI for limited assistance with grammar, spelling, clarity, organization, readability, or similar editorial purposes.

The author must retain control over the final language and meaning.

4.4 Accessibility and Formatting: AI may be used for accessibility-related functions, transcription, formatting, summarization, or other technical processes that do not replace the author’s creative contribution.

4.5 Translation: AI-assisted translation may be permitted when the author or translator meaningfully reviews, corrects, and assumes responsibility for the final translation.

Tundra Swan Press reserves the right to inquire and request disclosure of substantial AI involvement in translation.

5. Uses Requiring Disclosure

Authors must disclose material use of generative AI when AI has contributed to the creation or transformation of material submitted for publication. Disclosure is required when AI has been used to:

  • Generate substantial portions of prose, poetry, dialogue, or other literary text.
  • Substantially rewrite or transform the author’s original writing.
  • Generate or substantially alter artwork, illustrations, photographs, or other visual material.
  • Generate substantial portions of a translation.
  • Create characters, scenes, passages, poems, essays, or other creative elements that remain materially present in the published work.
  • Generate audio, video, voice, or other creative components.
  • Substantially restructure a manuscript beyond ordinary proofreading or copyediting.
  • Or in other ways, made a significant creative influence to the submitted work.

Disclosure should identify, in reasonable terms, what AI was used for and the extent of its involvement.

Authors are not required to disclose every minor interaction with an AI-enabled software feature.

6. Substantive AI-Generated Content

Tundra Swan Press distinguishes between AI-assisted authorship and AI-generated authorship.

AI-Assisted Work

A work may be considered AI-assisted when the human author remains the principal creative force and used AI in a supporting capacity. Examples include grammar and punctuation editing tools, brainstorming, limited language refinement, research assistance, or technical support.

AI-Generated Work

A work may be considered AI-generated when AI produces substantial creative material and the human contribution consists primarily of selecting, arranging, lightly editing, prompting, or approving that material.

Tundra Swan Press reserves the right to decline publication of work that does not meet its human-authorship standard, even when the use of AI has been disclosed.

Disclosure does not automatically make otherwise ineligible material acceptable for publication.

7. Author Disclosure Statement

When material AI use has been used in the work, Tundra Swan Press may request an AI-use statement.

The disclosure may be structured in a manner similar to the following:

AI Use Disclosure: Generative AI was used during the preparation of this work for [brief description of use, such as brainstorming, copyediting, translation, or image generation]. The author reviewed and takes responsibility for the final published material. For more substantial use, the Press may request additional information concerning:

  • The AI system or tool used.
  • The portions of the work affected.
  • The general nature of the prompts or instructions.
  • Revisions, drafts, handwritten notes the author used in preparing the work.
  • The extent of human revision.
  • Whether AI-generated material remains in the final work; and
  • Whether AI-generated images, audio, or other components are included.

Tundra Swan Press will seek only information reasonably necessary to evaluate authorship and compliance.

8. Author Responsibility

Regardless of whether AI was used, the submitting author remains responsible for the final work. Authors must not knowingly submit AI-generated material that:

  • Contains fabricated quotations, citations, sources, or references.
  • Presents false factual claims as verified information.
  • Infringes on another person’s copyright, trademark, privacy, publicity, or other rights.
  • Unlawfully reproduces another person’s work.
  • Impersonates or improperly reproduces a living person’s identity or voice.
  • Contains undisclosed plagiarism or unattributed borrowing, or
  • Violates applicable law or the press’s other publishing policies.

Authors should exercise caution when AI systems are used for historical, scientific, legal, medical, technical, or other factual material requiring reliable sourcing.

9. Authorship Verification

Tundra Swan Press does not regard AI-detection software as definitive proof of AI authorship. AI detectors can produce both false positives and false negatives. Consequently, a detector score, algorithmic classification, stylistic observation, or other automated signal will not, by itself, establish that an author violated this policy. When authorship concerns arise, the Press may use a combination of evidence, including:

  • The author’s disclosure.
  • Drafts and earlier versions of the manuscript.
  • Revision history or writing records voluntarily provided by the author.
  • Notes, research materials, outlines, or source documents.
  • Discussion with the author concerning the work’s development.
  • Consistency between the submitted work and the author’s established creative process.
  • Factual and source verification.
  • Examination of suspected AI-generated passages.
  • Available metadata or document history, where appropriate and lawfully obtained, and
  • AI-detection tools as one limited signal among multiple forms of evidence.

Tundra Swan Press’s objective is not to prove that a manuscript “sounds like AI.” The objective is to determine whether the author has complied with our authorship and disclosure requirements.

10. Verification Process

If Tundra Swan Press has a reasonable concern regarding undisclosed or potentially disqualifying AI use, it may initiate an authorship review.

Step 1: Editorial Concern

Tundra Swan Press will identify the nature of the concern and review the available manuscript and editorial evidence.

Step 2: Author Inquiry

TSP may contact the author and request clarification concerning the creation of the work and any AI tools used.

The author will generally be given a reasonable opportunity to explain the work’s development and provide supporting material.

Step 3: Additional Review

Where appropriate, TSP may review drafts, revision history, notes, source materials, or other relevant evidence voluntarily supplied by the author.

TSP may ask the author for live writing and/or editing exercises to validate the author’s writing style related to the work.

TSP will not require authors to disclose confidential information unrelated to the authorship question.

TSP may provide the work to an independent editorial review panel for professional expertise and discussion to reach a decision in the verification process.

Step 4: Editorial Determination

After considering the available evidence, the Press may determine that:

  1. The work satisfies the TSP’s authorship standards.
  2. The work satisfies the standards but requires an AI-use disclosure.
  3. Additional revision or clarification is required.
  4. The work does not satisfy TSP’s human-authorship requirements, or
  5. The available evidence is insufficient to decide in which case TSP may decline the submission without making a finding of misconduct.

11. Fairness and Presumption of Good Faith

Tundra Swan Press will approach authorship questions in good faith and will not presume misconduct solely because a manuscript has characteristics sometimes associated with AI-generated writing. TSP recognizes that:

  • Human writing styles vary substantially.
  • Authors may use different drafting and revision practices.
  • AI-detection systems are imperfect.
  • Heavily edited human writing can appear algorithmically generated.
  • AI-assisted writing can be difficult to distinguish reliably from entirely human writing.

Consequently, an author will have a meaningful opportunity to respond to a substantive authorship concern before TSP makes an adverse determination, except where circumstances make such an opportunity impracticable.

12. Consequences of Noncompliance

Responses will be proportionate to the circumstances and may include:

  • Requesting disclosure or clarification.
  • Requesting revision of AI-generated material.
  • Requiring an appropriate acknowledgment or disclosure.
  • Removing AI-generated material from a manuscript.
  • Rejecting a submission.
  • Withdrawing an accepted manuscript before publication.
  • Correcting or updating published work.
  • Suspending consideration of additional submissions from the author, or
  • Terminating a publishing agreement where permitted by that agreement.

Intent, materiality, the extent of undisclosed AI use, the author’s response, and the potential effect on the integrity of the publication may all be considered. An honest failure to understand the policy will generally be treated differently from deliberate concealment or misrepresentation.

13. Previously Published Work

Authors submitting previously published or substantially pre-existing material must disclose any material AI connection when required by this policy. Authors are also responsible for ensuring that they possess the necessary rights to submit the work to Tundra Swan Press. If substantial undisclosed AI-generated material is discovered after publication, the Press will assess the circumstances and may release a correction, clarification, withdrawal, or other appropriate resolution.

14. AI-Generated Images and Other Media

Where a publication includes AI-generated or substantially AI-altered visual, audio, or multimedia material, Tundra Swan Press may require disclosure and appropriate labeling.

TSP may decline material when its use of AI:

  • Creates unacceptable copyright or rights concerns.
  • Falsely represents a real person, event, location, or photograph.
  • Misleads readers about the nature or provenance of the material.
  • Conflicts with the press’s artistic or editorial standards, or
  • Creates legal or reputational risks that the press cannot reasonably manage.

Authors are responsible for complying with the terms applicable to the AI tools they use.

15. Privacy and Confidentiality

Authors should not submit confidential, unpublished, proprietary, personally identifiable, or otherwise sensitive material to an AI system unless they have an appropriate legal and practical basis to do so.

Tundra Swan Press will treat information supplied during an authorship review with reasonable discretion and will use it for legitimate editorial, contractual, legal, or compliance purposes.

TSP does not require authors to provide complete records of every AI interaction unless such information is reasonably necessary to resolve a material authorship question.

16. Editorial Use of AI

This policy applies primarily to material submitted by authors and contributors. Tundra Swan Press may itself use AI-assisted tools in editorial, administrative, accessibility, or production workflows.

Where editorial AI use materially changes an author’s creative work, Tundra Swan Press will seek to preserve the author’s substantive creative control and, where appropriate, communicate material changes to the author.

The Press remains responsible for editorial decisions made by its staff, contractors, or representatives.

17. No Reliance on AI Detection Alone

Tundra Swan Press expressly rejects a policy under which an automated AI detector determines authorship.

AI-detection scores may be considered as an investigative lead, but they are not treated as conclusive evidence.

If TSP feels an AI-detection program may provide insight regarding the use of AI-generated content, TSP will discuss the matter with the author. If an AI-detection program is used, TSP will only use a program such as Pangram that specifically outlines that all data is encrypted and stored securely on their servers to be displayed in user’s history. Pangram does not share or sell user data to third parties.

A manuscript will not be rejected solely because an AI detector assigns it a particular probability or percentage of AI authorship.

Human editorial judgment and the totality of relevant evidence remain controlling.

18. Policy Updates

AI technologies, industry practices, applicable laws, and publishing standards are evolving rapidly. Tundra Swan Press may periodically revise this policy.

Material changes will be communicated to authors and contributors as appropriate.

The version of the policy applicable to a submission will ordinarily be the version in effect when the manuscript is submitted unless the applicable publishing agreement provides otherwise.

19. Author Acknowledgment

By submitting work to Tundra Swan Press, the author represents that:

  1. The author has read and understands this policy.
  2. The author has disclosed material use of generative AI when required.
  3. The author has not knowingly misrepresented AI-generated material as wholly human-authored when such representation would violate this policy.
  4. The author has reviewed the final submission and accepts responsibility for its contents.
  5. The author will cooperate reasonably with legitimate authorship-verification inquiries.

Submission of a manuscript does not guarantee acceptance or publication.

20. Guiding Standard

This policy’s main question isn’t just whether AI is used during the process, but who is responsible for the creative work. Did AI play a role in creating it, and if so, was that role disclosed when it was appropriate to do so?

Tundra Swan Press embraces authors who thoughtfully apply innovative technologies, ensuring that published creative work retains meaningful human authorship, accountability, and editorial integrity.