WTWH content is created by human journalists, because there is no AI that can provide context, distill knowledge, and tailor data to our meet audiences’ extremely specific informational needs.

That said, we leverage productivity gains from AI to increase the throughput of content – including original reporting and research, data-based journalism, and authentic-voice content – that can only be produced by human editorial teams.

This framework for the use of AI is intended to enable our editorial teams to better serve our readers and audiences while avoiding or minimizing the risks posed by the irresponsible use of AI, including doubts about factual accuracy, the reputational risk that comes with passing off AI-generated work as our own, and the harm that AI-generated copy can cause on search engine rankings. This framework applies to anyone contributing journalism to a WTWH publication, including journalists on staff and freelancers.

• WTWH journalists embrace AI as a productivity tool. If AI can do a task at least as well as a human can, let it – but always with human involvement. A human always reviews, corrects, edits, and contextualizes content generated or edited by AI.
• In our news articles and other journalistic content, we do not publish verbatim text generated by AI; our writing is original and in our style.
• Information gleaned from AI will not be published unless verified as accurate by reporters, who should always be able to cite the original source of information included in an article.
• We use AI as a backend and ideation tool. These uses might include formulating interview questions, suggesting story ideas, sourcing stories, outlining stories, suggesting headlines, writing marketing emails, suggesting SEO tags and keywords, summarizing existing text that already has been published by WTWH or other organizations.
• We use AI to edit and improve content from freelancers as well as guest contributors who are not professional journalists.
• We use AI to edit and improve submitted press releases.
• We use AI to generate text for social-media posts and to draft other communications about WTWH content.
• We use productivity gains from AI to develop more content that can only be produced by human editorial teams. This includes original reporting, research, data-based journalism, and authentic-voice content.
• Reporters and editors are vigilant about the risks related to biases that AI-generated text may introduce, in order to avoid introducing that bias into our work.
• Editors are aware of the risks posed by AI and are vigilant about maintaining our AI policy; however, because it can be difficult or impossible to discern all the ways a reporter has used AI, ultimate accountability for the accuracy and originality of editorial work lies with each reporter.
• We do not enter sensitive, confidential or proprietary information into a generative AI tool such as ChatGPT. We respect our sources’ security and safeguard our own data and information regarding our business practices.
• We continue to evaluate and amend this policy as necessary.