Argentina’s leading digital media outlets are learning — the hard way — how to maintain credibility while integrating AI. Their wins and mistakes are a masterclass for any communication strategy.
One of the most repeated questions in the newsrooms I studied for my thesis was this: do we need to tell the audience we used AI? According to every journalist I interviewed, the answer was yes. But the reasons they gave revealed something more interesting than a simple yes or no.
This article is adapted from my postgraduate thesis, “Between the Human and the Automatic: Generative AI and the Role of Journalism as a Cultural Mediator” (Master’s in Communication, Culture and Media Discourses, 2026).
The consensus: audiences have a right to know
Every interviewee — without exception — agreed that if AI use has a significant impact on the content produced, it must be disclosed and explained to the audience. Not as a shameful confession, but as part of the trust contract.
“It would be unethical to sell the audience content as if it were produced by a human, backed by a reputable brand,” was how one editor put it. And that principle has direct implications for any organization that produces public-facing communication.
The cost of not being transparent
There were cases where outlets published AI-generated or AI-assisted content without disclosing it, and when it came to light, the reputational damage was disproportionate to the actual technical error. “There have already been cases of outlets coming out badly when they misused AI. It seriously damages the reputation of the brand and eventually of the journalist who makes the mistake,” noted a senior journalist at Clarín.
Audiences are also developing a growing sensitivity to patterns of automated writing. “There are certain style cues that are very obvious,” described one Infobae reporter. Readers are beginning to detect the robotic tone, the overly tidy structure, the absence of a real voice. And when they detect it without a disclaimer, they feel deceived.
The black box: the problem of algorithmic ppacity
There’s a deeper problem beyond transparency about whether AI was used: the opacity of how it works. Generative AI processes are, in many cases, what researchers call a “black box” — they produce outputs without it being possible to trace the reasoning behind them.
For journalism, this is a fundamental ethical problem: if the content can’t be audited, if there’s no traceability in the process, editorial responsibility dissolves. And communication without identifiable editorial accountability has a structural — not situational — credibility problem.
What brands can learn
The lesson from newsrooms is transferable to any organization that uses AI to communicate.
First, establish a clear policy: in which communication uses is AI involved, and how is that disclosed?
Second, don’t delegate judgment: using AI to help produce content doesn’t mean it should decide what to publish.
Third, understand that transparency doesn’t weaken you: audiences value honesty about tool use, as long as it’s accompanied by human oversight and accountability.
Credibility isn’t built by avoiding AI. It’s built by being clear about how you use it and what human role exists in that process.
*Image credit: Created with Nano Banana.
