The Transparency Trap
When AI Labels Backfire
Commentary by Jaci Clement
Originally appeared in The Latest, Fair Media Council’s weekly newsletter.
We’ve hit a strange moment in the relationship between the public and the media: The public says they want transparency — until they actually get it.
Case in point: Research shows that when an article is labeled “AI-generated,” people trust it less, even if the story is accurate.
That’s the transparency trap.
Relationships are always tricky, especially between the public and the media. But if you’ve looked at the headlines of late, you may have noticed things are a bit fluid. Traditions are being tossed aside, institutions are wobbling, and the very idea of public service is very much under renovation. Whether you view this as progress or collapse actually isn’t the point.
The part that captures our attention is this: If AI is now at work interpreting the world for us, will we someday look back and say these were the very best of times?
AI and Media Literacy
There are many issues at play, not the least of which is an atmosphere so immersed in media that people no longer notice it — it’s so much a part of our lives it’s on par with air and water. Now, if the public has not been given the time and training to discern AI-generated content — they haven’t — and the labels that the media are using are supposed to increase trust but fail, then what exactly comes next?
The thing is, AI can only work with what its operators know. If the operators aren’t media-literate… If they don’t understand context… If they can’t spot bias, omission, nonsense, or the now-infamous “AI slop”… Then AI becomes a very confident storyteller with no understanding of the truth — and a human pushes the button that says publish.
What does that do to the public square, and to its people?
Here’s another problem, and it’s one the conspiracy-minded will absolutely love: Today’s digital ecosystem isn’t designed to chronicle history. It’s designed to rewrite it. Algorithms reward novelty, emotion, and speed. Accuracy? Not part of the code. Relevancy? Optional. Fairness? Wrong platform.
Without strong guardrails put in place by humans, what we’re left with is a public wandering through that public square without direction. It’s the media that’s supposed to supply the map.
The Trust Gap
Some have forgotten and others have never known that the reason news exists is to help them live better lives, keep them safe, and protect their communities. But at this moment in history, all of what they do and how they do it is in flux, too.
And in this mess, news organizations may have their very best chance to rediscover their purpose. Much of the news industry has been so focused on reinvention that it’s willing to try anything. But does everything need reinvention? Re-examination? Sure. Updating. Rethinking. Modernizing for a world where information is fast, fractured, and frictionless makes sense — but it needs to be built upon that old-school foundation that insists on finding truth, relevance and something no AI can provide: meaning.
Transparency is good, but substance is better. Labels matter, but process matters more. AI won’t save journalism — but it may kill it if journalists don’t understand it.
History tells us something here: Newsrooms once handed over their content to the Internet for free. Then spent decades trying to get people to pay for it. The parallels could be a teaching moment. But that relies on people seeing the pattern. People, not AI.
The question facing the news industry isn’t whether to use AI. It’s if the people steering it know enough — about reporting, verification, fairness, and community — to use it to create a better breed of journalism. The kind that puts a story in perspective, doesn’t create false narratives, and doesn’t so handily convict someone in the court of public opinion in an age when anyone may be accused of anything.
And one more thing: When trust in the media is already fragile — as survey after survey shows — a label won’t rebuild it. Why would the public trust a label when they don’t trust the institution behind it? And institutions, after all, are made of people.
Also by Jaci Clement: FMC Fast Chat
Learn More: About Jaci Clement