AI Transparency

Working draft — not final legal text

This content is a working draft, not yet adopted. Missing details and statements requiring confirmation are listed below. It must be reviewed and completed before final publication.

Required before final publication

Insert and verify the final legal entity details, registered office, authorised representative and official contact before final publication.
Confirm this draft against the final entity structure and applicable law. This policy has not yet been adopted.
Confirm the actual human editorial responsibility and rules for institutional positions and publication-level notices. The suggested publication statement is example wording, not confirmation that each publication has been reviewed.
Confirm appropriate disclosures for existing AI-generated imagery and any future AI content or direct AI interaction.

Recommended general notice

“Some content published on or through the AIGC website may be generated, assisted, summarised, translated or otherwise processed using artificial intelligence systems. Such content may contain factual errors, omissions, biases or other inaccuracies. AI-assisted content is not necessarily an official position of AIGC unless expressly identified as such. Users should independently verify material facts and obtain appropriate professional advice before relying on such content.”

Human responsibility

Use of AI does not remove human responsibility for AIGC publications. Where AIGC publishes an institutional position, the relevant human review and approval process should remain controlling.

Official positions

AI-generated or AI-assisted text does not become an AIGC position merely because it appears on an AIGC channel. Official status should depend on the Council or other authorised approval route established by AIGC governance.

Publication-level notices

Where AI materially contributes to the generation or transformation of a public-facing publication, AIGC may attach a short publication-level transparency statement. The wording should describe the role of AI accurately rather than applying the same label to every use.

Suggested publication statement

“This publication was prepared with the assistance of artificial intelligence tools. The final content was reviewed under human editorial responsibility. Any institutional position attributed to AIGC reflects the applicable AIGC review and approval process.”

Images, audio and video

Synthetic or materially AI-generated images, audio or video should be labelled where required by applicable law or where disclosure is appropriate to avoid misleading the audience. Deepfake or materially manipulated content should receive particularly clear disclosure.

Direct AI interaction

If AIGC later provides a chatbot, AI tutor, automated assistant or other direct AI interaction, users should be informed that they are interacting with an AI system unless that fact is already obvious from the context.

Public-interest text and EU access

Where AIGC publishes AI-generated or AI-manipulated text intended to inform the public on matters of public interest and the relevant EU AI Act requirements apply, AIGC should follow the applicable transparency rules, including any human-review/editorial-responsibility conditions and required disclosures.

Machine-readable provenance

Where technically practical and relevant, AIGC should preserve available provenance or metadata provided by AI systems and publishing tools, particularly for synthetic media. This policy can be expanded when AIGC develops a more formal content provenance framework.

Contact AIGC