Audiences already expect newsrooms are using generative AI, and they want to know about it. A 2025 University of Minnesota survey found that 52 percent of 1,128 respondents said it's "very important" that newsrooms disclose AI use in article text, even when humans verified the information. A 2024 Reuters survey of participants from six countries produced similar results, with up to half of respondents wanting disclosure.
Whether your newsroom uses AI or not, publishing a public AI policy can shift audiences from skepticism toward openness about your process. Many outlets have already done this.
Start with your rationale
Some policies open with an introduction explaining the thinking behind AI use. The Guardian published principles stating it only uses AI when it improves work quality, such as helping journalists interrogate data or suggesting corrections. NPR's ethics guidelines include a generative AI section that calls the technology a "great tool for some tasks that are 'hard to do but easy to verify,' such as helping identify key themes in large documents."
AP News includes information about its collaboration with OpenAI. Kristin Heitmann, senior vice president and chief revenue officer, said "AP firmly supports a framework that will ensure intellectual property is protected and content creators are fairly compensated for their work. News organizations must have a seat at the table to ensure this happens."
Drawing the line on yes versus no
A 2024 post from the nonprofit Trusting News found that "explaining exactly what the AI tool did, why it was used, and how humans were still involved helps demystify the use of the tool." One way to frame this is by listing when AI is acceptable and when it is not.
Some outlets state any AI-generated or AI-assisted content will be labeled as such (ProPublica, The Guardian, AfroLA). Others permit AI for text summaries, social media posts, or transcribing audio (Núcleo). Language translation is permitted at AP News, with the caveat that translated text will be edited before it reaches audiences.
On the negative side, Texas Tribune says it will not publish AI-generated text without a "rigorous verification and editing process." A 2026 study from suggests this boosts credibility. MIT Technology Review promises writers "do not use generative AI to draft articles or parts of articles." Many outlets state they won't use AI-generated or manipulated images. ElTOQUE specifies that sources' personal identities will not be uploaded to public AI platforms.
The technology is evolving quickly, and some newsrooms make clear their policies may change. The Narwhal notes that its AI working group "will regularly evaluate and update" standards. Texas Tribune uses the term "cautious experimentation" to describe its approach, emphasizing that journalists "must be critical of the tools."
No AI is a choice worth explaining
If your newsroom decides not to use generative AI, communicating that decision still matters. The Philly Plain Dealer has a "No AI" statement with quotes from its reporters explaining why they oppose the use of AI in their work.
AfroLA acknowledges in its policy that AI can be a "disruptive tool" but aims to "take every possible action so that it is used to enhance, rather than replace, human work." Pittsburgh's Public Source writes that its policy "emphasizes our values of human-centered journalism, transparency, accuracy, and privacy protection."
For journalists looking for guidance on where to start, AI for Writers provides practical coverage of how the technology intersects with editorial work. The resources from Poynter's AI ethics guidelines for newsrooms offer a deeper starting point.
Why this matters for writers
Writers are the ones who will have to follow the policy and explain its reasoning to sources, editors, and readers. If you are a writer drafting or reviewing an AI policy for your newsroom, the clearest signal you can send is whether AI is used for drafts, summaries, or transcriptions and where current decisions are deferred. Without that clarity, you leave room for assumptions that erode the trust the policy was built to protect.
Your membership also unlocks:
