Fake RNZ news about Hayley Holt circulating on the hijacked Facebook page of Trevor Mallard. Photo: RNZ Mediawatch
Newstalk ZB's Mike Hosking ended his show one day in August up in arms about scammers on Facebook.
They were using his name and image to sell junk or hoodwink people into giving away their financial information.
"This is all rubbish. It's spam. It's nonsense," he said.
Some of the ads in question said Hosking had "confirmed allegations".
One said his producer had been forced to cut to commercials after he "dropped a bombshell".
He was right. The ads are annoying. And false.
But as his producer Glenn Hart admitted, they can be alluring to a vulnerable audience.
"Certainly draws me in," he said.
"The one that says 'Mike drops bombshell, producers forced to cut straight to commercial', that is definitely plausible."
We can all get sucked in by a bogus headline, particularly if it's tailor-made to our tastes.
Hosking is far from the only celebrity on the receiving end of these scam ads.
They've proliferated across Facebook for months.
Many of them mimic the appearance of RNZ stories, plastering the broadcaster's name and logo above articles about well-known people doing something unspecified, but horrible.
One delivers the disturbing news that "Miriama Kamo's unexpected announcement has shocked the public".
Another shows 1News presenter Daniel Faitaua bruised and battered
.
If the headline is to be believed (which it isn't), the beating appears to have been administered after a live cross gone wrong.
There are hundreds more ads like this.
Politicians are another common target.
Several ads report the death of Chris Hipkins - though on the bright side for the very-much-alive Labour leader, they also name him as one of our "most beloved New Zealanders" in defiance of his preferred prime minister polling.
A fake scam ad falsely announcing the Leader of the Opposition's death. Photo: RNZ Mediawatch
Many of these are sent from real Facebook accounts that have been hijacked.
The official Facebook page of former Speaker of the House and current High Commissioner to Ireland Trevor Mallard has been used by people in Vietnam, Bangladesh and a host of other countries to run more than 250 ads - including a bogus RNZ article about TVNZ's Hayley Holt.
If you're wondering why these fake ads have been allowed to mushroom without any apparent intervention, you're not alone.
People have been complaining about the tsunami of literal fake news for months.
Dylan Reeve, a freelance writer and editor, has catalogued hundreds of the ads and reported them to Facebook's parent company, Meta.
A lot of the time he gets the same answer.
"Almost universally it's 'thanks for reporting it, we're not taking any action'," he tells Mediawatch.
The responses are generally boilerplate. They tend to arrive after exactly seven days. Reeve suspects the ads are getting to the end of a queue and then falling off without anyone taking action.
He says Meta has the resources to better moderate these ads, but it chooses not to employ enough real people to do the work.
"Facebook could hire someone with the ability to spot a fake ad and spot it before it gets run. They say they do an approval process before ads get placed but what that is is a mystery."
"The ones that really shock me are the ones that include AI videos of politicians. I've got videos of Chris Luxon, Chris Hipkins, Winston Peters. I've got AI videos of Ashley Bloomfield talking about health products."
But how is Facebook getting away with simply saying the ads don't break its guidelines or indeed its terms of service?
They would appear to breach copyright law at the least, along with New Zealand trading laws.
The ones showing figures like Bloomfield spruiking bogus health products could even be defamatory.
When Reeve approached the Ministry of Business, Innovation and Employment about the issue, he got back an anodyne statement essentially saying it is aware of scams and trying to prosecute those committing fraud.
But it's one thing to play whackamole with scammers in southeast Asia, and another to go after the giant tech company that profits from their business while making lax attempts at moderation.
Paul Thompson, the chief executive of RNZ, says he's weighing up his options for action.
"They're ripping people off. People who are spending time reading them are being exposed to scams. Their data is being harvested. They're ripping media companies off as well - for RNZ they're using our brand and our people."
Despite that, Thompson doesn't feel like media companies currently have the resources to go up against Facebook's army of lawyers.
He wants stronger government regulation aimed at making the tech companies that host these ads culpable for the damage they do.
"The incentives aren't there for it to be combatted.
"Local regulation would help. But I think it's also a global problem that needs a multinational response"
In the meantime, it's a tricky situation for our news organisations.
They are low on resources to fight this kind of thing, partly because Meta has hoovered up most of their ad money.
Now the company that took so much of their business appears to be gaining even more money from scammers actively eroding what trust they've retained.
It's a double whammy, a vicious cycle - and one that could be ramping up as AI makes these sorts of ads easier to mass produce.
But it's not all bad. As Hosking found out, AI is a double-edged sword. It might be used to generate fake ads using your name and likeness, but at least it can also be used to generate other stuff, like catchy country songs about the nation's most popular breakfast news host.
Hopefully the trade-off is worth it.