AI-Generated Death Hoaxes Hit New Zealand Athletes and Families

A wave of AI-generated fake articles is hitting New Zealand’s biggest sports names, and the most recent target is Warriors NRL head coach Andrew Webster. His wife Emma was the subject of a fabricated death notice this week, and 1News reports the post is part of a pattern that has hit All Blacks, Crusaders and Chiefs figures in the same week. Emma Webster, the woman he credited on stage when he won the 2023 Dally M Coach of the Year, is alive.

Webster, 44, took the question at a Warriors media conference this week, and Crusaders fullback Johnny McNicholl told 1News his biggest worry is older relatives believing the cancer claims. Netball’s Tactix and other ANZ Premiership sides have begun issuing their own warnings to fans.

The Names Behind the Fake Death Notices

Webster is the most visible name drawn into the wave so far, but he is far from the only one. 1News listed former and current All Blacks Will Jordan, David Havili and Brad Weber among recent targets, with Weber falsely reported to have stage four cancer. Crusaders fullback Johnny McNicholl said he had seen a post that morning claiming his teammate Macca Springer had been terminated for misbehaviour.

• A fabricated death notice for Andrew Webster’s wife Emma, built from a real photo of the couple at the 2023 Dally M Awards
• A false stage four cancer diagnosis for All Blacks halfback Brad Weber
• A claim that Crusaders back Macca Springer had been “terminated his contract for misbehaviour”
• An entirely made-up Chiefs player called “Jack Pugh,” with a photo lifted from an American football player of the same name who died in April

Some of the posts target people who do not exist at all. The “Jack Pugh” case shows how thin the template is. 1News reported a Chiefs player of that name had died, with a photograph lifted from an American football player of the same name who died in April. The same template was re-used for a Warriors version, with the imagery tweaked to fit each club.

Where the Posts Come From

Most of the pages pushing these stories are not run from New Zealand. A separate fact-check tracing a Warriors-themed page to operators in Vietnam found the page posing as a fan account for the club, posting around the clock and linking out to ad-stuffed external sites in the comments.

That investigation found the same template adapted for other sports. One fabricated post claimed a 35-year-old Warriors employee had been fatally shot breaking up a bar fight, illustrated with an AI-generated image of a man in the club’s jersey. Another claimed Warriors star Roger Tuivasa-Sheck had donated $500,000 to the family of a five-year-old girl allegedly killed in Alice Springs in April 2026. The fact-check found no media reports of the shooting and the club had made no such announcement, and there was no evidence of the donation or the statement attributed to Tuivasa-Sheck.

None of those claims were real. A Google Image search of several of the photos surfaced a Made with Google AI watermark embedded in the pixels, the digital fingerprint of an image generator. Closer to home, a 1News investigation in February documented at least 10 New Zealand-themed Facebook pages that took real press releases and police notifications, ran them through an AI to rewrite them, and re-published them with synthetic images. In one case, a still photograph of a teenager killed in the Mount Maunganui landslide was edited to show her dancing, and in several posts the raw AI prompts were left visible, exposing the production line.

The Cost on the Receiving End

For Webster, the cost of a single post is a question he was asked to answer in public. RNZ reproduced the fuller exchange at the Warriors media opportunity, with the 44-year-old coach saying he had stayed in his lane by leaving the policing of fake images to “smarter people than me.”

To hear you talk like that about my wife is not very nice. But look, I’m not on social media; I only hear about it through people like you today, but at the same time if you want to live in that world and be a part of it, you’re gonna get this fake news.

Webster, the 44-year-old Warriors head coach, said the exchange at the media conference this week caught him off guard.

McNicholl, the Crusaders fullback, said the posts he was most worried about were the ones read by relatives. “Especially with my parents’ generation,” he told 1News, “they’re definitely believing some of those cruel statements about people getting cancer.” The fake cancer diagnosis given to Brad Weber is exactly the kind of post McNicholl described: a real name, a real club, and a medical verdict made up out of whole cloth. Players do not always see the posts themselves, he added, but the damage is done in the family group chat before anyone at the club hears about them.

Tactix general manager Kate Agnew told 1News the wider problem goes beyond any one team. “Some people will say, ‘well any publicity is good publicity’,” she said. “But in this case we are concerned that this is putting out a really negative environment and an unhealthy rhetoric that people are engaging with under false pretences.” The Tactix and other ANZ Premiership sides have begun issuing direct warnings to fans, urging them not to share posts from unverified pages.

The pattern crosses codes, and crosses countries of origin. A 1News February investigation found a separate wave of AI-driven news pages rewriting real local press releases with synthetic images and posting them on Facebook, and the National Emergency Management Agency has had to issue its own warnings about AI imagery during emergencies. Once every post is suspect, real reporting has to climb over a wall of doubt, which is the cost the fake pages are selling.

Two Motives, One Click

Why does anyone bother? Massey University marketing professor Bodo Lang offered two reasons to 1News, neither flattering. The first is the influencer chase. Tragedy, in that model, is just another niche, with the post as a hook and the comments as the call to action. Lang sketched the calculus in plain terms: a big enough audience, a tight enough niche, and a steady stream of emotionally charged content to keep the algorithm pushing the page.

“For most users, it would probably be to become an influencer of some sort, to have a following that’s, let’s say, more than a million,” Lang told 1News. “We all know the stories of Mr Beast and Joe Rogan, you know, their living is online now. So I think there are many, many users who wish to emulate that and create a following in whatever niche they want to carve out.”

The second motive is older and uglier: malware. Lang told 1News that some operators “just want to plant malicious software on your computer, so their goal is not so much true engagement, but it’s actually just you clicking on that link.” If 10,000 people click and the payload freezes a hard drive or extracts personal information, “that’s the mission they accomplished for them,” he said. He urged readers not to click through to external sites on any of these posts unless they were posted by a verified source.

The mechanic in both cases is the same: an emotionally loaded headline, a famous face, and a call to action in the comments. The same delivery chain has been used to push fake AI tool install pages past the real ones in search results, with one campaign using a one-line PowerShell payload to harvest developer sessions, Slack tokens, VPN configs and crypto wallets from anyone copying the command.

Whack-a-Mole at the Platform Level

New Zealand Rugby told 1News it was working through a process with Meta, the company that owns Facebook, where most of the posts are surfacing. Tactix and other ANZ Premiership sides have started putting out their own warnings. Neither response has so far stopped the flow.

Lang described the process of reporting and removing these pages as a game of “whack-a-mole.” “As soon as you report a page, it’s just going to reappear somewhere else really, really quickly,” he said. The February 1News investigation made the same point from a different angle: one of the worst offenders, a page called NZ News Hub, vanished from Facebook a day after 1News began asking questions about it, and it was not clear whether the operators or Meta had removed it. Within days, near-identical pages were back.

The Trust Erosion Nobody Can Plug

The longer-term damage is to the news audience, not the targets. AUT research, cited in 1News’s February investigation, found that only 32% of New Zealanders trust the news.

The National Emergency Management Agency has had to issue its own warnings about AI-generated imagery during severe weather events, including the deadly Mount Maunganui landslide. Gisborne District Council and Tairāwhiti Civil Defence put out a joint statement warning of fake pages pretending to be news outlets and sharing AI-generated images about local events and emergencies. The combined effect is a public sphere in which any single image, including the ones in this story, has to be checked before it is believed.

The same fight is now landing in the courts. A UK member of parliament has filed a High Court claim over deepfake design choices, testing whether AI developers can be held liable for what their tools help produce. New claimants are joining, the filing shows. The case will set a precedent for what model design decisions, not just specific outputs, can be challenged in court.

The numbers behind the wave

• 32% of New Zealanders trust the news, per the most recent AUT Trust in News survey cited by 1News
• 10 New Zealand-themed Facebook pages documented in 1News’s February investigation, all of which used AI to rewrite real press releases and pair them with synthetic images
• 209 posts reviewed on one of the worst offenders, NZ News Hub, in January 2026 alone, none of them with AI imagery labelled as such
• 10,000 clicks is the threshold at which Lang says a malware campaign would consider its mission accomplished

Frequently Asked Questions

Who has been targeted by the latest wave of fake AI articles in New Zealand?

Warriors NRL head coach Andrew Webster, whose wife Emma was the subject of a fake death notice, is the most visible name so far. All Blacks Will Jordan, David Havili and Brad Weber have all been hit, with Weber falsely reported to have stage four cancer. Crusaders fullback Johnny McNicholl has spoken about seeing fabricated posts naming his teammates, including one falsely claiming teammate Macca Springer had been sacked for misbehaviour.

Who is making these posts?

A fact-check tracing a Warriors-themed page, “Warrior Spirit NZ,” to operators in Vietnam, not New Zealand, was the most concrete attribution so far. A 1News investigation in February 2026 documented at least 10 New Zealand-themed Facebook pages that scrape local press releases and police notifications, rewrite them with AI, and pair the rewritten text with synthetic images.

Why are people making them?

Massey University marketing professor Bodo Lang told 1News there are two motives. The first is to build a social media following large enough to monetise, treating tragedy as a content niche. The second is to use the post as a delivery mechanism for malware, with the goal of getting users to click through to a third-party site that can extract personal information or freeze their hard drive.

What is being done about it?

New Zealand Rugby said it was working through a process with Meta, which owns Facebook. ANZ Premiership netball sides, including the Tactix, have begun issuing their own warnings to fans. Researchers and Lang both describe the enforcement process as a “whack-a-mole” cycle in which reported pages are quickly replaced by new ones.

How can a reader tell a fake AI article from a real one?

Researchers at AUT and Victoria University of Wellington say the most reliable defence is source verification, checking whether the post comes from a recognised news outlet rather than a similarly named page. Visual clues such as incorrect uniforms, gear that does not match local standards, or distorted text in images can help, but those cues are getting harder to spot as the underlying models improve. Source checking, not visual cues, is the main defence that still holds up.