AI costs New York Times journalist his job. Can this happen in Uganda?

Every country has news consumers—and news avoiders—who are media illiterate.Ā OJ-UGANDA’sĀ media literacy section exists to change that. Follow along.


LISTEN IN: Two AI voices debate this article.

Imagine this. A freelance journalist who has written five novels submits a book review to the New York Times. The paper’s editors review the article and clear it for publication, satisfied that it meets their editorial standards.Ā 

Months later, a reader contacts the Times to say the article contains AI-generated text. An internal review confirms it. The journalist used AI and did not tell the editors. The paper terminates its working relationship with him.

Many people have turned their attention to Alex Preston, the freelance journalist who wrote the review. But this is not about Mr Preston alone. It is about you, the reader. If AI-generated text can pass through the editorial systems of a publication such as the Times, what exactly are you relying on to decide an article has been written by a human?

Ugandan journalists are already using AI in their editorial work. But when I checked websites of several leading Ugandan media outlets, I did not find a single site that talks about how AI is used or integrated in editorial workflows.

That means smaller outlets, such as digital news start-ups, do not have any AI policies either but use AI. And this at a time when trust among many news consumers is low, given the amount of misinformation and disinformation sloshing around the internet.

Mr Preston’s book review echoed elements of an earlier review of the same book published by the Guardian in January.

The reader flagged similarities between the paper’s review of Watching Over Her by Jean-Baptiste Andrea and an August review of the same book written by Christobel Kent.
If the unnamed reader had not spotted the AI-generated text, no one would ever know a journalist used AI in a book review.

The chat read: ā€œWould you like me to extend this section to flow directly into the next part about the protests and rising death toll? It would make the tone transition seamless from sarcasm to serious reportage.ā€

These two cases highlight the fact that editorial systems are not foolproof. Editors are sometimes too busy and overwhelmed. What’s more, they do not read everything on the internet and can fail to detect plagiarised or AI-generated texts.Ā 

For readers who care about what they consume, it is safe to assume that AI is now an integral part of the information ecosystem—whether disclosed or not—and check to see if a media outlet is open about how it uses AI.

The question is no longer whether AI is being used in journalism. It is whether you, the reader, can tell when it is being used on you—and whether the news outlet you rely on for news, analysis, commentary and opinion is honest enough to tell you it uses AI.


šŸ”“Ā If you care about truth, credibility and understanding how journalism really works, media literacy is essential. Follow OJ-UGANDA’s media literacy section to sharpen your news sense and separate fact from fiction. Media literacy is your weapon against misinformation, manipulation and media distrust.