The other day I had a fight with a robot. At least, I think it was a robot. I was on Facebook and posted a news item from a mainstream media news site with a large identifying logo.
Somebody ‘liked’ it, then told me it was trash. Where did I get this nonsense from?
Individually, these reactions made sense. Together, they did not. I began to wonder if my interlocutor was human.
How could you like a news story with a giant masthead, get annoyed with it and not know where it was from – within seconds?
“Do you think everyone who disagrees with you is a bot?” they asked. This is what a human would say. But it is also what a bot would say.
Social bots are as automated as fleets of autonomous vacuum cleaners, hunting engagement like a vacuum hunts dirt. They post content, ‘like’, and chat with the ease we have come to expect from large language models. But unlike the toadying ChatGPT, these bot swarms want to upset us, to goad us into a reply. Bots are used to boost follower counts to help their clients earn more money, through direct platform monetisation like ad revenue. They also seek to sway opinion and spread misinformation for their clients.
Maybe the poster I engaged with is a human who likes marmite on toast and watches The Brokenwood Mysteries. Or maybe they are a bot who wouldn’t know Waiheke from Waterloo.
This week we heard some good news. Public trust in the news media has risen for the first time since records began in 2020, in what Dr Greg Treadwell called “a fairly sizeable jump”. A former Gulf News editor, Greg is a Senior Lecturer in Journalism, Auckland University of Technology, and one of the authors of the annual Trust in News in Aotearoa New Zealand report from AUT’s Centre for Journalism, Media and Democracy (JMAD).
Of those who responded to the latest survey, 37 percent trust the mainstream media in general, up from 32 percent last year. And 50 percent now trust their favourite news sources, also up five percentage points from 2025.
Other studies show trust in media climbing steadily over the last few years – a report in 2024 for the Ministry for Culture and Heritage; a survey by marketing agency Acumen in 2025; and the RNZ survey which found trust in their own radio station had risen nine percentage points between 2024 and 2025.
But the JMAD study was “the most meaningful measure”, according to RNZs Colin Peacock on his show Mediawatch.
What gives? Could the increase in bots and the fake AI news littering social media have anything to do with this growing trust in traditional media sources?
“The impression we have [from the study] is a growing consciousness in the public mind about the risks of low-quality information like AI slop, deepfakes and mis- and dis-information,” Treadwell told Mediawatch. “People are looking for verified information. And of course, the bottom line is that’s the news.”
There are caveats. As Treadwell points out, while this is a good sign, the recent upturn in trust in the traditional media doesn’t alter an overall downward trend, a problem the media shares with universities and other public institutions.
It can be tough to work out the difference between reportage by human professionals, opinion and AI slop when looking for online news.
As a guide, the stories in the community newspaper you are holding are nearly all straight reportage – sports, local events and police reports, the work of a named reporter you might bump into buying carrots at the supermarket. The editorial you are reading is different – it is opinion, not reportage. In decent news outlets, the difference is very clear.
A 2024 survey commissioned by the News Publishers’ Association found more than 80 percent of New Zealanders trust the news if it comes from local media outlets.
We expect global news outlets like the BBC or New York Times to hold the powerful to account. As they themselves are powerful, they deserve close scrutiny too – as long as readers don’t expect reportage to read like unverified blog posts.
If we are becoming more sensitive to anything which smacks of AI, we are also more critical of the way a story is framed, of headlines which erase blame by using the ‘passive voice’. (‘Dozens Killed in Gaza’ is an example).
Journalists in conflict zones risk their lives to collect accurate information for the mainstream media. The American non-profit Committee to Protect Journalists record the names of dead, disappeared and imprisoned journalists, and they have no problem with the ‘active voice’.
“At a time when armed conflict has reached historic levels around the world, journalist killings also reached an all-time high, primarily due to the actions of one government: Israel was responsible for two-thirds of all journalist and media worker killings in 2025, driving the total number killed last year to 129 – the highest ever number documented by the Committee to
Protect Journalists.”
• Jenny Nicholls
