AI & journalism: Progress or compromise?

 

Clawed, amused and instantly judgmental — that’s us. Give us a story, a situation, anything fresh and we are ready to dissect it with full force. That is exactly what happened when a recent incident at Dawn, Pakistan’s leading English daily, set off a wave of amusement, criticism and reflection across the country. A business story titled “Auto sales rev up in October” included more than the reporter intended: the newspaper accidentally printed the internal AI editing prompt used to refine the piece. Screenshots spread fast, opinions spread even faster and within minutes, social media had turned it into the latest national spectacle. Hilarious memes took over timelines and the public’s favourite pastime — collective judgment — kicked in effortlessly.

Yet beneath the laughter was something far more telling. A reputable, decades-old newspaper had slipped and with that mistake came serious questions about journalistic ethics and the rapidly shifting landscape of newsrooms. People began asking what many in the industry have been quietly debating: How much influence should artificial intelligence have in journalism? Where exactly do we draw the line between assistance and authorship? And does honesty in reporting lose its meaning when machines shape the words we present as our own?

The dilemma is deeper than AI-generated paragraphs or misplaced prompts. It lies in the fine line between keeping control and surrendering it. Tools like ChatGPT have made it easy to write, edit organize and process information at lightning speed. They have made workloads lighter, deadlines less suffocating and newsroom pressures marginally easier to manage. But convenience is not the same as integrity. So the question becomes: Do we need AI to think for us? And if we allow it to think, to what extent?

We live in an age that celebrates technological advancement — flawless rendering, impeccable algorithms, machine precision that often outshines human ability. Newsrooms around the world are not only acknowledging this shift but also adapting to it. Several international publications have introduced clear policies outlining how AI can be used to supplement reporting: generating summaries, cleaning grammar, analyzing data or creating outlines. These guidelines are designed to ensure that the final work remains human, that the voice stays authentic and that accountability rests on the writer, not the machine.

But not everyone is ready to embrace this blended future. Some media houses are outright rejecting AI, preferring to rely solely on human instinct, judgment and creativity. Their stance is simple: AI may produce beauty, but it is not human. It cannot feel discomfort, courage, conviction or the subtle pulse beneath a story. It cannot understand the moral weight of truth-telling or the cultural nuance of a region where a single sentence can shift public sentiment. Its efficiency cannot replace empathy.

And so, the tension persists. As journalists, many of us feel caught in an evolving puzzle. Those who want to preserve their voice — their human tone, their instinctive storytelling — now find themselves measured against the perfected cadence of an AI-generated paragraph. There is a growing, unspoken insecurity: Is the human word still good enough? Or is the industry slowly nudging us toward a future where the machine’s version sounds cleaner, safer and more “publishable”? Only time will answer that.

What we know for now is that journalism stands at a crossroads. We can continue to celebrate AI as a tool, not a crutch. We can continue to refine, not replace. We can continue to uphold the honesty that forms the backbone of this profession. But that requires transparency, discipline and a willingness to admit that while AI can assist, it cannot — and should not — own the narrative.

The Dawn incident, in all its humor and chaos, was more than a viral moment. It was a gentle reminder for educators, writers, editors and readers alike. The tools we rely on may evolve, but the responsibility remains the same. Human eyes, human errors, human judgment — these cannot be automated. And perhaps that is the real lesson here: technology can support journalism, but it cannot define it. Not yet. And not if we remain committed to safeguarding the soul of the craft.

—The writer is contributing columnist, based in Islamabad.

([email protected])

 

Get Alerts