New York Times fires contributor over AI-powered plagiarism
Artificial intelligence is already part of everyday life for a lot of people, from playlist suggestions to automatic document summaries. In journalism, though, things are quite different. The industry’s editorial standards demand originality, rigorous reporting, and accountability for every word published, which puts careless AI use on a direct collision course with the profession’s core values.
And that clash is exactly what came to light when the New York Times found itself at the center of an embarrassing case involving AI-generated plagiarism. The newspaper itself called the incident a serious violation of its editorial standards.
Freelance journalist Alex Preston, who had been contributing to the Times since 2021 and had written six reviews for the paper, was let go after readers noticed that a book review published on January 6 had passages strikingly similar to a piece from the Guardian that had been published roughly four months earlier. The review in question covered the book Watching Over Her by Jean-Baptiste Andrea.
The internal investigation confirmed what everyone suspected: the text had been written with the help of AI, and excerpts from the Guardian article were incorporated without any attribution. Preston later apologized to the Guardian, the New York Times, and the author of the original review, and stated during the inquiry that he had not used AI for the other six reviews he wrote for the paper.
The incident raises questions that go far beyond a single journalist or a single publication, and it shows that journalism is still trying to figure out how to coexist with this technology without sacrificing what matters most: credibility. 🔍
What actually happened
The story started to come into sharper focus when an attentive New York Times reader spotted very specific similarities between the review published by the paper and a piece the Guardian had previously published about the same book. These were not just similar ideas that could be chalked up to two people sharing the same opinion about a work. They were sentence structures, word choices, and even sequences of arguments that repeated from one text to the other — something hard to explain as mere coincidence.
This kind of similarity is exactly what plagiarism detection tools catch with ease, and it was what grabbed the public’s attention before any formal investigation even began. Following the tip from that reader, the Times launched an internal review of the published content.
The findings were unmistakable: Alex Preston had used artificial intelligence to write the review and, in the process, ended up including — in a way he himself described as unintentional — passages that belonged to the analysis previously published by the Guardian. The core problem was not exactly the use of the tool itself, but the way it was used: without proper oversight, without critical review of the generated content, and without any effort to ensure the final result was genuinely original.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Language models like those behind AI writing tools are trained on massive volumes of text pulled from the internet, and there is a real risk that they reproduce — whether inadvertently or not — structures and content from previously published texts. When the professional fails to apply that filter, the outcome can be exactly what happened here.
A New York Times spokesperson publicly acknowledged the issue. According to the Guardian, the newspaper’s representative stated that the author’s reliance on AI and the use of unattributed work from another journalist constituted a clear violation of Times standards. The contributor was let go, and the paper sent an unambiguous signal that editorial integrity is non-negotiable, regardless of pressure for productivity or the temptation to take technological shortcuts.
The episode exposed a vulnerability that many newsrooms around the world still don’t quite know how to address. 📰
The role of AI in journalism and where the danger lies
It would be unfair to say that artificial intelligence has no place in journalism. Many newsrooms already use automated tools for specific tasks like transcribing interviews, analyzing large volumes of data, identifying patterns in lengthy documents, and even generating drafts for low-complexity factual news stories such as sports scores or standardized financial reports. In these contexts, AI works as a support tool that frees up the journalist to focus on what truly requires human judgment: interpreting the facts, choosing the angle of the story, and taking ethical responsibility for what goes to press.
The problem starts when that logic is flipped and AI takes center stage in the creative process, especially in formats that depend on an authorial voice, like reviews, columns, and analysis pieces.
Large-scale language models, known as large language models, work by predicting the most likely sequence of words based on patterns learned during training. This means they are, by nature, systems that recombine existing information — not engines of genuine creation. When a journalist uses this kind of tool to write an opinion piece or cultural analysis and does not critically review what was generated, they are essentially handing authorship over to a system that has no editorial accountability, no firsthand knowledge of the facts, and no way to guarantee the output is original.
The New York Times case is a practical and quite costly example of how this can go wrong, especially when the resulting text gets too close to something already published by another outlet.
Another point worth paying attention to is the issue of transparency. Much of the conversation around AI in journalism still revolves around whether outlets should disclose when they use these tools in content production. Some have already adopted clear disclosure policies, while others are still defining their internal guidelines. What the Times episode makes clear is that a lack of transparency, combined with improper use of the technology, creates a double risk: publishing content that is not original and damaging the reputation an entire newsroom built over decades because of one poorly calibrated individual decision.
And that applies to any outlet, no matter its size. 🤖
The numbers that show how big this problem really is
To understand why this particular case carries so much weight, it helps to look at the broader data on the use of artificial intelligence in newsrooms around the world. The numbers help put the scale of the challenge the industry faces into perspective.
According to researchers at the University of Maryland, in 2025, just under 10% of news content from newspapers already contained AI-generated text. That might seem small at first glance, but we are talking about a significant slice of the daily news cycle being produced with some degree of automation — often without the reader having any indication of it.
On the media company side, the picture is even more revealing. A survey conducted by the International News Media Association (INMA) found that a staggering 97% of publishers are investing in AI for a variety of purposes across their operations. This ranges from headline optimization and content distribution to administrative task automation and audience analysis.
At the same time, public perception does not necessarily match that corporate enthusiasm. A Pew Research Center survey showed that roughly half of Americans view artificial intelligence as something problematic for journalism. There is a legitimate concern that the quality and reliability of information could be compromised when machines take on tasks that have traditionally depended on human judgment.
This combination — massive adoption by newsrooms and significant distrust from the public — creates a complex landscape. And as the technology behind these language models keeps improving, the trend is that spotting misuse will become increasingly difficult, both for readers and for editors themselves. 📊
Editorial standards under fire
Journalism’s editorial standards exist for a very concrete reason: to ensure the public can trust what it is reading. Originality, proper attribution of sources, fact-checking, and accountability for published content are not just bureaucratic rules. They are the pillars that set professional journalism apart from any other kind of text production. When a professional cuts corners around those pillars — whether by using AI irresponsibly or simply by not reviewing what they submit — the damage is not just personal. The credibility of the entire outlet is at stake, and rebuilding that trust after a public plagiarism scandal is a long and costly process.
What makes this case especially significant is that it happened at one of the most respected and influential newspapers in the world. The New York Times is not some unknown publication trying to grow through shortcuts. It is an institution with over 170 years of history, recognized globally for the quality and rigor of its journalistic coverage. That does not make it immune to mistakes, as this case proved, but it puts the incident under a much larger microscope than it would be in another context.
The backlash was significant precisely because the expected standard is higher, and any deviation becomes more visible. For the rest of the industry, the episode serves as a clear warning that no newsroom is automatically protected from the risks that come with careless adoption of new technologies.
From a practical standpoint, what many newsrooms still need to define more clearly are internal policies on the use of artificial intelligence in content production. This includes points like:
- Establishing at which stages of the editorial process AI can be used
- Defining which types of content are appropriate for this kind of assistance
- Creating clear protocols for reviewing material generated by automated tools
- Determining whether use of the tool needs to be disclosed to readers
- Implementing more robust plagiarism checks before publication
Without these guidelines, each journalist ends up making individual decisions that may or may not align with the outlet’s values, and the result can be exactly the kind of situation the Times faced. Technology moves fast, but editorial governance structures need to keep up. ⚠️
What we take away from this story
At the end of the day, what this episode shows with great clarity is that the debate over artificial intelligence in journalism is far from settled. The technology will keep evolving, the tools will become increasingly sophisticated and accessible, and the pressure for productivity in newsrooms is not going away. But all of that needs to be balanced with responsibility and a real understanding of the risks involved, especially when it comes to content that depends on an authorial voice, critical analysis, and genuine originality.
The case of Alex Preston at the New York Times is not an outlier that can be dismissed as an exception. It is a symptom of a tension present in newsrooms all over the world — between the push for speed and volume on one side, and the need to maintain rigorous editorial standards on the other. Finding that balance is not simple, and there is no magic formula, but ignoring the issue is clearly not a viable option.
Plagiarism, whether intentional or the result of negligent AI use, has real consequences for the professional’s career and the outlet’s reputation. Preston lost a years-long collaboration with one of the biggest newspapers on the planet, and the Times had to publicly deal with a stain on its editorial credibility.
The good news — if you can even call it that — is that newsrooms have the power to put more rigorous safeguards in place to prevent this kind of situation. Plagiarism detection tools, clear AI usage policies, more robust editing processes, and an editorial culture that values transparency are all concrete paths to reducing risk.
The lesson here is simple, even if it is not easy to put into practice: tools are tools, and human judgment is still irreplaceable when the subject is editorial integrity. Artificial intelligence can be a powerful ally in journalism, but only when used with discernment, oversight, and full awareness of its limitations. Without that, a shortcut can quickly turn into a dead end. 🎯
