Trust is handmade by people. It cannot simply be manufactured by machines.
At the Bangladesh Journalism Conference 2026, two veteran journalists with decades of experience across print and digital media offered a quiet but urgent reminder: the greatest danger artificial intelligence poses to journalism is not that it will replace reporters, but that it will make it easier to get things wrong. In an era when public trust in news is already fragile and misinformation spreads faster than correction, the question is not whether newsrooms will use AI, but whether they will use it with the discipline and honesty that journalism has always demanded of itself. The technology is neither savior nor villain — the character of the people and institutions wielding it will determine which it becomes.
- AI-generated misinformation is not a looming threat — deepfakes and manipulated content are already distorting elections and political crises around the world right now.
- The real danger inside newsrooms is not the algorithm but the journalist who stops checking facts, trusting a machine to do the work that only human judgment can do.
- Smaller newsrooms in developing countries face a particular pressure point: limited budgets and scarce technical expertise make it tempting to adopt AI tools without the ethical frameworks to govern them.
- Entry-level journalism jobs built on routine, structured reporting are quietly disappearing, narrowing the pipeline through which the next generation of journalists has always learned the craft.
- The industry is still drawing the line between AI assistance and AI authorship — and until that line is clear, transparency with audiences is the only currency newsrooms can spend to protect their credibility.
At the Bangladesh Journalism Conference 2026, two veteran journalists — Michael Cooke, former editor of Canada's Toronto Star, and Murdoch Davis, with fifty years across print and digital media — gathered not to debate whether AI would transform journalism, but to ask how newsrooms could survive that transformation with their credibility intact.
Their answer was less about technology than about character. Any serious AI policy, they argued, must be built on the foundations journalism has always rested on: truth, accuracy, and honesty. For smaller newsrooms in places like Bangladesh, where resources are thin, this does not require reinventing the wheel — it requires studying what larger outlets have built, adapting those frameworks, and beginning the conversation seriously. The important thing, they said, is simply to begin.
The clearest line they drew was around verification. AI should never replace the human work of confirming what is true. Careless use of the technology increases the risk of factual errors, and factual errors erode the trust that audiences extend to news organizations — trust that, once lost, is nearly impossible to recover. On transparency, their position was equally firm: for now, audiences should always know when AI has been involved in reporting or editing. The public is already skeptical, and newsrooms cannot afford to deepen that suspicion.
When pressed on which risk concerned them most, their answer was grounded: the biggest threat is not AI itself, but sloppy journalism. Weak editorial judgment and careless reporting have always spread misinformation — AI simply accelerates the damage. In politically polarized societies, that acceleration is especially dangerous. AI-generated deepfakes and manipulated content are already distorting elections globally. This is not a future problem to prepare for. It is a present one to reckon with.
Entry-level jobs, they acknowledged, are genuinely vulnerable. Routine tasks — summarizing official statements, processing structured information — can now be handled by machines. Some of those opportunities will disappear. But fieldwork, source cultivation, verification, and editorial judgment remain irreducibly human. Those cannot be automated away.
Looking ahead, Cooke and Davis offered a simple but demanding forecast: the future of journalism depends almost entirely on the choices newsrooms make about how to behave. The same tools that could help honest journalists work more efficiently could also supercharge low-quality content operations and disinformation networks. Trust, they concluded, is not something a machine can manufacture. It is built by people, one story at a time, through years of getting things right.
Two veteran journalists sat down during the Bangladesh Journalism Conference 2026 to talk about something that has begun to reshape every newsroom on earth: artificial intelligence. Michael Cooke, who spent years leading Canada's Toronto Star, and Murdoch Davis, a fifty-year veteran of print and digital journalism, were there to discuss not whether AI would change the news business—it already has—but how newsrooms should handle it without destroying the thing that makes journalism matter in the first place: the trust of the people who read it.
The conversation began with a straightforward question: what are the minimum ethical standards a newsroom should follow when using AI? The answer was equally straightforward. News organizations need to build serious AI policies now, not later. But here is what matters: the foundation of any such policy cannot be about the technology itself. It has to be about truth, honesty, and accuracy. For smaller newsrooms in places like Bangladesh, where budgets are tight and technical expertise is scarce, the path forward does not require starting from scratch. These organizations can study what larger outlets have already built, adapt those frameworks to their own circumstances, and begin the conversation seriously. The important thing is to begin.
When the discussion turned to how smaller newsrooms could use AI responsibly without compromising their credibility, the answer revealed something crucial: you do not need a large budget or advanced infrastructure to do this right. Basic internal guidelines work. The key is deciding what AI should never do—and verification and factual reporting are at the top of that list. AI should never replace the human work of checking facts and confirming what is true. The real threat, Cooke and Davis explained, is not the technology itself. It is whether journalism gets things right or wrong. Careless use of AI increases the chances of factual mistakes, and factual mistakes damage the trust that audiences place in news organizations. That trust, once broken, is extraordinarily difficult to rebuild.
The question of transparency came next. Should audiences always know when AI has been used in reporting or editing? For now, yes. The public is already skeptical about AI and journalism, and news organizations cannot afford to do anything that weakens credibility further. But the boundaries are still being drawn. Is spell-checking AI-assisted editing? What about grammar correction or shortening a story? The industry is still figuring out where those lines should be drawn. What is clear is that the conversation needs to happen openly, not behind closed doors.
When asked which risk concerned them most—whether AI would amplify misinformation or whether it would eliminate entry-level journalism jobs—the answer was direct: the biggest concern is simple journalism getting things wrong. Sloppy reporting, weak editorial judgment, and careless use of technology can all accelerate the spread of false information. In politically charged environments, those mistakes can spread with terrifying speed and become even more dangerous. Entry-level jobs are indeed vulnerable. Routine tasks built around official statements, meeting minutes, and structured information—work that junior reporters have traditionally handled—can now be processed and summarized by AI quickly. Some entry-level opportunities will likely disappear. But real reporting, fieldwork, verification, and editorial judgment still require human journalists. That work cannot be automated away.
The conversation then turned to a question specific to South Asia: could AI-generated misinformation become even more dangerous during elections or political crises in polarized societies? The answer was sobering. It is not a future risk. It is already happening. Deepfakes, manipulated videos, and AI-generated misinformation are spreading globally right now. Polarized societies are especially vulnerable during elections and political crises. The speed and scale of AI-generated misinformation make it extraordinarily difficult for journalists to counter false narratives before they spread widely online. This is not a theoretical problem. It is a present one.
Interestingly, when asked what mistakes newsrooms are making in their rush to integrate AI, the answer was that many newsrooms are not rushing at all. Many organizations are moving cautiously, and some are moving too slowly. But the real mistake would be to wait until AI tools are everywhere before having serious discussions about ethics, transparency, and newsroom standards. Those conversations need to happen now. Looking ahead five years, the future of journalism depends almost entirely on how news organizations choose to behave. Newsrooms genuinely committed to truth, fairness, and honesty will continue upholding those values regardless of technological change. But disinformation websites and low-quality content operations could also expand rapidly using the same tools. Trust is not manufactured by machines. It is built by people, one story at a time, through years of getting things right.
Citas Notables
The core issue is not technology itself. The core issue is whether journalism gets things right or wrong.— Michael Cooke and Murdoch Davis
Newsrooms genuinely committed to truth, fairness and honesty would continue trying to uphold those values regardless of technological changes.— Michael Cooke and Murdoch Davis
La Conversación del Hearth Otra perspectiva de la historia
You mentioned that the core issue isn't technology itself, but whether journalism gets things right or wrong. That feels almost obvious, but why do you think newsrooms are struggling with it?
Because AI makes it easy to move fast without thinking. You can generate a summary in seconds, but you haven't verified it. You can process information at scale, but you haven't checked sources. Speed and scale are seductive. They feel like progress. But they're not progress if the story is wrong.
So the risk isn't that AI will replace journalists. It's that AI will replace the thinking part of journalism.
Exactly. A junior reporter doing basic reporting—covering a meeting, summarizing a statement—that work can be automated. But the moment you automate it, you've lost the human who might have asked a follow-up question, who might have noticed something didn't add up. That's where mistakes happen.
In a polarized country like Bangladesh, where misinformation spreads fast, how do you even fight back against AI-generated false information?
You can't fight it after it's already spread. You have to prevent it from spreading in the first place. That means journalists have to be so credible, so clearly committed to truth, that audiences trust them more than they trust a viral video. It's about building a reputation over years, not defending it after it's damaged.
You said smaller newsrooms don't need to build policies from scratch. What does that actually look like in practice?
It means sitting down and writing down: we will not use AI for verification. We will not use AI to replace reporting. We will use it for efficiency—organizing data, drafting summaries—but always with a human checking the work. It doesn't require a lawyer or a consultant. It requires honesty about what you're doing and why.
Is there a version of the future where AI actually strengthens journalism?
Yes. If newsrooms use it to do more reporting, to reach more sources, to verify more thoroughly. But that requires discipline. It requires saying no to speed when accuracy is at stake. Most organizations aren't built that way right now.