Artificial Intelligence Policy

Artificial Intelligence (AI) Policy

Frontier Advances in Applied Science and Engineering (FAASE) • Tinta Emas Institute

1) Purpose

This expert-authored policy sets clear, actionable rules for the responsible use of AI in submissions and editorial workflows. It preserves research integrity, transparency, reproducibility, privacy, and intellectual property—while allowing carefully disclosed assistive use.

2) Scope & Core Principles

  • Applies to authors, editors, and reviewers.
  • Human accountability is non-delegable; AI is not an author or reviewer.
  • AI use must be disclosed, limited, auditable, and lawful/ethical.
  • No undisclosed generation, alteration, or concealment of research content.

3) Definitions

  • Generative AI: Systems that create text, code, images, audio, video, or synthetic data.
  • Assisted use: Language polishing or formatting with no change to scientific meaning.
  • Substantive use: Idea generation, drafting, literature summarising, coding, image editing, data synthesis, or analytical steps.
  • Public tools: Web services that may store prompts/outputs, risking confidentiality.

4) Authors — Responsibilities

  • Verify all AI-assisted text, code, figures, and data. Check facts, citations, and calculations.
  • Retain full control over study design, analysis plans, and scientific interpretation.
  • Respect licences for third-party materials/models; document provenance.
  • Do not upload confidential manuscripts, personal data, or proprietary datasets to public AI tools.

5) Acceptable vs Prohibited Use

Permitted (with disclosure) Prohibited (or require prior ethics/legal clearance)
Language polishing; grammar/spelling; neutral formatting (e.g., references style) without altering meaning. Fabricating/altering data, spectra, or images; concealing artefacts; “in-painting” results.
Outline suggestions or readability improvements where authors write and verify the final text. Undisclosed drafting of substantial sections or literature reviews.
Code refactoring hints followed by human review, testing, and documentation. Uploading confidential manuscripts, personal/sensitive data, or proprietary code to public tools.
Assisted figure layout (not data content) with disclosure and availability of originals. Auto-generating citations or using fabricated/misattributed references.

6) Mandatory AI Disclosure

Declare any substantive AI use at submission, within the manuscript (Methods/Acknowledgements), and in the cover letter. State the tool, provider, model/version, access date, purpose, extent, and verification steps. Generic statements are insufficient.

Item What to state
Tool & provider e.g., “ChatGPT (OpenAI), GPT-4.1; accessed 11 Oct 2025”
Purpose & stage Language polishing; outline; code suggestion; figure layout; etc.
Extent Sections affected; approximate proportion of text/analysis
Human verification Fact-check, recomputation, unit tests, image forensics, replication steps
Data sensitivity Confirmation that no confidential/personal data were entered into public tools
Example: “We used ChatGPT (OpenAI, GPT-4.1; accessed 11 Oct 2025) for language polishing of the Introduction and outline suggestions for Section 3. All content was verified and edited by the authors. All data, analyses, and conclusions are the authors’ own.”

7) Data, Images, and Code

  • Data: Label simulated/synthetic data; justify use; provide generation parameters/seeds; never mix with empirical data without clear separation and rationale.
  • Images/Figures: Disclose AI-based enhancement (denoising, upscaling). Provide originals on request. No fabrication or concealment.
  • Code: Provide runnable code, pinned dependencies, and tests. Authors are responsible for security and licensing.

8) Editorial & Reviewer Use

  • Do not paste confidential manuscripts or peer reports into public AI systems.
  • AI may assist with grammar or checklists; all judgements and reports must be authored and owned by humans.
  • Submission confidentiality and data protection are mandatory at all times.

9) Compliance, Audits, and Sanctions

  • Undisclosed or inappropriate AI use may lead to rejection, correction, retraction, or notification to institutions/funders.
  • FAASE may request raw data, original images, code, prompts, and outputs for verification.
  • Corrections, expressions of concern, and retractions follow established editorial ethics procedures.

10) One-Minute Compliance Checklist

  • I disclosed tool, provider, model/version, access date, purpose, extent, and verification.
  • I verified facts, citations, numbers, code, figures, and data. No fabricated references.
  • No confidential/personal/proprietary content was uploaded to public AI tools.
  • All scientific claims and interpretations are authored, owned, and justified by humans.
  • Original images/data/code can be provided on request; licences are respected.

11) Contact

Editorial Office — https://jurnal.tintaemas.id/index.php/faase/index (About / Contact).