AI Usage & Transparency Policy

AI may assist the work. Humans remain responsible for the journalism.

Editors and contributors are accountable for what INYIM publishes regardless of which software helped organize, draft, format, research or produce the work.

AI tools may be used as assistants for tasks such as:

• research leads and background organization
• transcription and interview-note organization
• headline, SEO and metadata ideation
• copy editing, structure and readability suggestions
• formatting, coding, debugging and publishing workflows
• image resizing, restoration planning, caption checks and other production support
• creative ideation for clearly illustrative, non-documentary material

Use of an AI tool does not make its output a verified source.

• AI output is not used as evidence for a factual claim without independent verification.
• Quotes, interviews, documents, events or sources may not be fabricated.
• AI-generated text should not be auto-published without human review.
• AI must not be used to impersonate a source or falsely represent direct access.
• Automated output does not replace editorial judgment on fairness, sourcing, context or corrections.

Claims introduced through an AI-assisted workflow are checked against appropriate primary or reliable secondary sources before publication. When an AI system provides a link, quotation, date or citation, the underlying source is opened and verified rather than trusted from the AI response alone.

Routine editorial adjustments such as crop, exposure, color balance, resizing and faithful restoration are distinct from inventing visual facts.

Synthetic or generative illustration may be used for clearly illustrative or creative purposes when appropriate. When the use of AI-generated imagery is material to how a reasonable reader would understand the image, it should be labeled or otherwise made clear.

Routine utility assistance — such as grammar suggestions, formatting, code help, transcription or headline brainstorming — does not normally require a story-level disclosure because the published material remains human reviewed and sourced.

Disclosure is appropriate when generative AI materially creates content a reader could reasonably mistake for original human reporting, documentary evidence or an authentic photograph, recording or statement.

Confidential, embargoed, legally sensitive or source-identifying material should not be entered into an AI service unless the tool and account are approved for that level of information and the editorial team has considered the privacy and contractual implications.

When in doubt, the sensitive material stays out of the tool.

AI may assist with website code, debugging, CMS operations, structured data, accessibility checks, performance work and other technical tasks. Those changes are subject to human review and normal publishing safeguards before they are intentionally made live.

This policy applies to published output and editorial practice; it is not a promise that every internal software interaction will be publicly logged.