TheInternationalists.org permits carefully reviewed use of artificial intelligence to support its editorial work on international theatre, cross-cultural collaboration, and performance across borders. This policy sets standards for that use. Human contributors and editors remain responsible for published material, including its accuracy, attribution, cultural context, and suitability for readers.
AI may assist with outlining, organising research notes, suggesting questions, improving clarity, preparing draft summaries, and identifying material for further investigation. It may also support technical tasks such as formatting structured information or developing website features. Permission to use a tool does not establish that its output is accurate or ready for publication.
Verification is required. AI-generated statements must be checked against sources appropriate to the subject. Names, production credits, dates, venues, quotations, translations, and references require particular care. A source suggested by an AI system must be located and examined before it is cited. A plausible title, link, or passage is not sufficient evidence that a publication exists or supports a claim.
AI output is not a substitute for documentary evidence. Articles should cite the underlying programme, archive, publication, interview, institutional record, or other relevant source. Editors must distinguish information actually found in a source from an inference or interpretation. Where evidence is unavailable or contradictory, that limitation should remain visible.
Historical records must retain their integrity. AI must not be used to invent missing production details, reconstruct an undocumented quotation, or generate supposed correspondence, reviews, interviews, or archival discoveries. A hypothetical example may help explain a practice, but it must be labelled clearly and must not be attributed to a real person or production.
These standards also apply to coverage of the earlier Internationalists directors’ collective. Generated text cannot establish organisational continuity, membership, endorsement, or a current partnership. Such claims require evidence independent of an AI system’s response.
Translation requires contextual review. AI may help prepare a working translation or identify passages needing further attention. Wording that affects an article’s interpretation should be checked by a person competent in the relevant language and subject. Dialect, humour, verse, culturally specific references, and multilingual dialogue call for particular scrutiny.
Machine-assisted translations must not be presented as an artist’s authorised wording without evidence of that authorisation. Quotations translated for an article should identify the published translation used or make clear that the translation was prepared for the article. Where reliable review is unavailable, material should not be published as a verified translation.
AI-generated media must be distinguishable from documentation. Generated images may be used for clearly identified conceptual illustrations. They must not be presented as photographs of actual performances, historical venues, costumes, programmes, or rehearsal sessions. Generated or materially altered representations of identifiable people require editorial review and appropriate permission where relevant. Synthetic voices or likenesses must not imply an artist’s participation or endorsement.
Disclosure should explain meaningful AI involvement. When generated text, translation, imagery, audio, or interpretation forms a substantive part of published material, an accompanying note should identify its role and the review undertaken. Routine spelling checks or formatting assistance generally do not require an article-level notice. A disclosure must accurately describe the work; it must not imply verification that did not occur.
Editors should examine AI-assisted material for cultural stereotyping, omitted perspectives, inappropriate generalisations, and inaccurate treatment of names or identities. Language and cultural practices should be described with attention to their specific contexts. Claims of firsthand experience, attendance, interviews, or testing may appear only when that work actually took place.
Confidential and restricted material requires protection. Contributors must not submit private correspondence, unpublished scripts, embargoed material, personal information, or restricted production documents to an AI service without appropriate permission and review of the service’s relevant settings and terms. Access to material does not automatically authorise its use in an AI workflow.
Time-sensitive information, including festival dates, residency eligibility, fees, and application deadlines, must be checked with the relevant organiser or official source. Where website tools use generated output, their limitations should be explained. Estimated results must remain distinguishable from verified facts and tested calculations.
Contributors should disclose substantive AI assistance to the editor when submitting work. An AI system does not receive a human byline or replace accountable authorship. Material that cannot meet these standards should be revised, withheld, or removed.
Readers may report suspected fabricated references, misleading media, translation problems, or other concerns through the Contact page. Please include the affected URL and supporting details. Confirmed errors should be addressed under the site’s correction standards. This policy may be revised as editorial practices and available technologies change.
Further reading: UNESCO’s Recommendation on the Ethics of Artificial Intelligence, including its principles on transparency, privacy, fairness, and human oversight. UNESCO