By Prof. (Dr.) Pooja Rana Professor, Mass Communication & Head — PhD programme (Social Sciences), Amity University Haryana
Journalism has somehow managed to survive many disruptions. It survived the arrival of radio, which was speculated to takeover newspapers. It survived television, which was expected to replace radio. It survived the internet, which was assumed to smash all three. However, instead of losing itself, the profession kept evolving. What is unfolding now with artificial intelligence, belongs to that same family of disruption, but there is a difference this time, and it is not a small one. It is moving faster than ever before, and is intensifying into the actual reporting. Which brings me to a question I never thought of facing before: can content not written by a human being ever earn the kind of trust journalism has spent over a century building?
I have been watching this profession for about thirty years now; first as a student trying to understand why some messages persuade while others fall flat, later as a researcher, and today as someone who spends hours trying to prepare young minds for newsrooms that barely resemble the ones I once knew. The question resists a binary answer, and I would regard any attempt to force one with considerable skepticism. To answer it, we first need to go back to what trust in journalism was actually built on, and be equally honest about what AI can and cannot yet offer.
Trust Is a Social Contract, not a Writing Style
In communication theory, we often speak of the source-credibility model, the fairly old but still useful idea that audiences never judge a message purely on its own terms. They judge it in relation to the one who is delivering it. A well-written sentence, on its own, does not carry much weight. It carries weight because someone stands behind it, someone the reader believes has an expertise, has verified the facts, and can be held accountable if they have not.
This is precisely what makes AI-generated news a different category of problem altogether. AI can generate a clean, fluent, well-organised, and even persuasive text. But journalism was never about fluency, it was always about accountability. When a journalist’s byline appears on a story, it carries an implicit promise: a trained professional is standing behind every line, and can claim it. AI cannot make that promise, not yet. It has no professional identity to protect, no reputation to lose, and no capacity to be questioned in the way a human reporter can be.
Automation Without Accountability Is a Structural Risk
Newsrooms everywhere, including plenty here in India, are already using AI to draft routine copy, summarise long documents, and put together quick bulletins. Used responsibly, these aren’t threats. They are just an extension of a process that has always involved tools, beginning from the telegraph to the wire service to the word processor.
The risk begins the moment automation moves from assisting the editorial process to replacing it. A newsroom that allows AI-generated content to reach the public without any human review is not just adopting new technology; it is altering the structure of accountability that journalism has always depended on. This is exactly why conversations on AI regulation, model transparency, and content labelling currently underway around the globe are not peripheral policy matters. They go to the heart of journalism’s core purpose. How these conversations are resolved will determine whether public can continue to trust what is placed in front of them every morning.
What Judgement Requires That Automation Does Not Yet Offer
There is a real gap between processing information and exercising judgement. AI systems are exceptionally good at first. They can analyse enormous volumes of data, detect patterns, and generate text at a speed no human reporter can ever match. What they cannot yet do is the kind of contextual, often unspoken judgement that experienced journalists develop almost by instinct: understanding what a fact means within a particular social or political context, recognising when a source is holding something back, and deciding what deserves prominence and what does not.
This is not a gap that will close with the next software update. It points to something more basic than a technical limitation, the difference between being competent and being responsible. A system can be competent at generating text. Responsibility, in the professional and ethical sense that journalism requires, is something only a human being can hold.
The Greater Danger Lies in Content That Appears Credible
If you ask me what worries me most, it is not clumsy AI copy. Most readers can spot that easily and dismiss it. The bigger concern is the content that is polished, well argued, and reads exactly like professional journalism, yet has bypassed the verification processes that give journalism its integrity. When content appears credible, audiences rarely interrogate its origin. This is precisely the condition under which misinformation thrives in, not through content that looks suspicious, but through content that looks entirely legitimate.
This is the reason transparency cannot be optional. News organisations that use AI in content production have a professional obligation to disclose it to their audiences. The absence of disclosure, even when unintentional, erodes the very trust journalism is supposed to be built on.
What This Means for the Profession Going Forward
I am not arguing that newsrooms should push AI out. That position strikes me as unrealistic and, even undesirable. Used responsibly, these tools can strengthen journalism considerably, freeing trained professionals from repetitive tasks to focus on investigation, analysis, and storytelling that requires genuine human insight.
What the profession must insist upon, without any compromise, is that human editorial judgement remains the final authority before anything reaches the public. AI may draft, summarise, and even assist, but it must not decide. This principle needs to be written into newsroom policy, taught explicitly in journalism education, and treated as a professional standard rather than a matter of individual discretion.
As educators, we carry a particular responsibility here. The students entering our classrooms today will be running newsrooms within the next decade. Our job is neither to train them to avoid these tools nor to adopt them uncritically, but to understand with precision where their usefulness ends and where professional judgement must take over.
A Position I Hold with Confidence
Public trust in journalism has never rested on how quickly a story is disseminated. It has always rested on the understanding that a real, accountable professional stood behind it. Artificial intelligence can help us produce news faster and more efficiently than ever before. What it cannot yet do is stand behind that news the way a trained journalist does.
Until that changes, the newsroom cannot afford to lose its journalists, regardless of how advanced the tools sitting on their desks become.
Prof. (Dr.) Pooja Rana is a Professor of Mass Communication at Amity University Haryana, with over two decades of experience in media education and research. Currently, she heads PhD programme (Social Sciences) at Amity University Haryana.
