AI and automation policy.

A person is accountable

(Re)in Asia may use artificial intelligence and automation to support parts of its publishing workflow. Every published item remains the responsibility of a named journalist or accountable editor. AI output is not treated as a source and is never, by itself, evidence that a claim is true.

How tools may assist

Subject to editorial review and appropriate data safeguards, tools may help with:

• Transcribing interviews and converting audio to searchable notes.
• Translation, with important quotations and technical terms checked against the source language.
• Summarising documents for research, while reporters read and verify the underlying material.
• Organising data, identifying patterns, checking calculations, or generating code used in analysis.
• Grammar, style, headline options, metadata, tagging, and other production tasks.
• Monitoring public sources or flagging possible story leads for a journalist to assess.

What AI may not do

• Publish a story without meaningful human review and approval.
• Invent or reconstruct quotations, sources, events, documents, data, or visual evidence.
• Impersonate a source, conceal synthetic communication, or misrepresent how reporting was obtained.
• Make a final editorial decision about newsworthiness, fairness, corrections, or allegations of wrongdoing.
• Turn confidential source material, personal data, or unpublished commercially sensitive information over to a system that is not approved for that use.
• Rewrite another publisher’s work in place of original reporting.

Verification and responsibility

The journalist and editor must verify every material fact against reliable sources, whether or not a tool was involved. Names, figures, quotations, citations, and links receive particular scrutiny because generative systems can produce plausible but false output.

Use of a tool does not lower our standards or shift responsibility to a software provider. Errors are handled under our Corrections and Complaints policy.

Source protection, copyright, and personal data

Staff must use approved tools and follow our privacy and security requirements. Confidential source information, embargoed documents, subscriber data, and personal data should not be entered into public AI systems unless an editor has approved the tool and the specific use.

We respect copyright, licensing, and attribution. AI-assisted work must add genuine editorial value and may not be used to disguise copying or evade a publisher’s terms.