AI and Journalism: The Real Question Isn't About Jobs

Earlier this year I came across a LinkedIn post by Mia Bockenhoff that made me stop. She was writing about AI in journalism, and instead of the usual panic about job losses, she made a different argument.

Her point: "The most effective approach to ensuring the credibility, dependability, and transparency of journalistic work is to combine the rationality and efficiency of AI with the empathy and diversity of humans."

That is the right framing. It just doesn't go far enough.

‍

The job question is a distraction.

Every industry having an AI conversation right now is having the wrong one. The question is not "how many jobs does AI take?" The question is: how do you design AI systems that people actually trust?

Journalism makes this concrete. When AI writes a news summary, readers need to know: is this accurate? Is it complete? Who verified it? If the answer is unclear, the output is worthless, regardless of how efficiently it was produced.

Efficiency without credibility is not progress.

‍

The real opportunity.

Journalism has a credibility problem that predates AI. Opinion dressed as fact. Speed prioritised over accuracy. Commercial interests shaping editorial decisions. These are human failures, not AI failures.

AI, used well, could fix some of this. If AI handles the mechanical work of structuring, summarising, and cross-referencing information, journalists are freed to do the hard work: source verification, ethical judgment, the reporting that requires real human contact.

AI output is only as good as the data it is trained on. Human output is only as good as the system it operates in. Both need to be honest about their limits.

‍

What trust actually requires.

I study trust in AI. Not as an abstract concept, but in the design of systems that real people interact with every day. What I have found is that trust is not built by capability alone. It is built by transparency, consistency, and honesty about what a system can and cannot do.

Journalism needs the same thing. An AI-assisted newsroom that is transparent about how its tools work, where its data comes from, and where human judgment overrides the algorithm, that is a more trustworthy newsroom than one that pretends the old model still works.

The profession has a genuine opportunity to rebuild credibility by being honest about how it is changing. Most industries just hide the seams.

‍

The broader point.

Journalism is not a special case. Healthcare, government, retail, public spaces, every domain where AI is now operating carries the same question: how do you deploy a system that people will actually accept and use?

The answer is not better marketing. It is honest design. Build transparently. Explain what the system does. Do not ask people to trust what they cannot see.

That applies to a news algorithm. It applies to a kiosk in a hospital corridor. The principle is identical.

‍

----------------------------

Justin Dean is the director of Envent Australia and is currently completing his PhD research in AI trust and human-centred design at the University of Newcastle.

‍

‍