CapKet’s AI and Automation Policy covers software and artificial intelligence used to assist newsroom work. The publisher remains responsible for what is published; use of a tool is not proof of accuracy or human editorial review.
Permitted assistance
Tools may assist with transcription, document organization, translation drafts, data cleaning, code, headline alternatives and research leads. Outputs are treated as unverified until checked against reliable sources.
Verification and review status
Material claims in new reporting must be checked against primary evidence before publication. Unverified claims must be omitted or the article held as a draft. Automated word-count, similarity, formatting and link checks are screening tools, not fact-checking or an independent human review. Some existing articles are AI-assisted and explicitly state that no independent human review is claimed. A named reviewer may be credited only for work they actually performed; the site must not imply that every article has received independent human review.
Disclosure
Material use of generative systems will be disclosed when a reasonable reader would want to know—for example, a substantially synthetic illustration, translated interview or data analysis central to the story.
Images and identity
Synthetic visuals must be clearly labelled and must not misleadingly depict a real event or person. We do not use AI to imitate another publication’s design, fabricate documentary evidence or remove ownership information.
Confidentiality and copyright
Confidential source material, unpublished market-sensitive information and personal data must not be entered into external AI tools without an approved basis. Staff must review training, licensing and output terms before using a tool in production.
Errors
AI-assisted errors are corrected under the same Corrections Policy as any other error. Send concerns to info@capket.com.