Photo: Getty Images
Photo: Getty Images

The BulletinJune 9, 2025

AI job losses are coming – and entry-level workers could be the first to go

Contributing writer
Photo: Getty Images
Photo: Getty Images
newsletter

After years of hype and hesitation, artificial intelligence is beginning to displace real jobs, writes Catherine McGregor in today’s extract from The Bulletin.

To receive The Bulletin in full each weekday, sign up here.

The quiet start of a loud disruption

For many workers, the presence of artificial intelligence in the office has so far been subtle –  a bit of text from ChatGPT here, an AI-generated image there. But a profound shift in the labour market is coming, many business and AI insiders say. Anthropic, maker of the Claude AI models, has found that while AI is being used mainly for the augmentation of tasks, a transition toward automation – actually doing the job – is well underway. “Most [workers] are unaware that this is about to happen,” Anthropic CEO Dario Amodei told Axios. “It sounds crazy, and people just don’t believe it.”

His warning was echoed by Axios co-founders Jim VandeHei and Mike Allen, who said they had “talked to scores of CEOs” about how they’re thinking about AI. “Every single one of them is working furiously to figure out when and how agents or other AI technology can displace human workers at scale,” they wrote. “The second these technologies can operate at a human efficacy level, which could be six months to several years from now, companies will shift from humans to machines.”

Tech jobs may be the canary in the coal mine

If that sounds alarmist, the numbers suggest otherwise. A global survey of senior executives found that 10% of all job roles could be eliminated by AI within five years. That figure, described as “modest” by Dr Paul Henderson in his recent Maxim Institute report, Gone for Good: AI and the Future of Work, would still cause “a marked disruption in New Zealanders’ lives”.

Some sectors are already showing signs of strain. Software development, in particular, has seen entry-level jobs disappear with the rise of coding assistants and no-code platforms. The IT news site CIO reports that teams are being restructured around AI tools, favouring experienced developers who can supervise machine-generated code over junior staff who once cut their teeth writing it. As Meta’s Mark Zuckerberg put it earlier this year, mid-level engineers – and presumably the junior coders aspiring to join them – may become unnecessary “in 2025”.

A grim forecast for new graduates

It seems that those searching for their first office jobs are most at risk. Amodei has predicted AI could wipe out half of all entry-level white collar jobs in the next five years and in the New York Times (paywalled), Kevin Roose reports that numerous firms are “making rapid progress toward automating entry-level work”, while others are becoming “AI-first”, evaluating whether a job even needs a human before posting the vacancy.

While this trend is most visible in the US, its implications are global – including for New Zealand, where many graduate roles mirror their overseas counterparts. The IMF has warned that up to 60% of jobs in advanced economies may be affected by AI. If so, the dream of a secure professional career may become a luxury afforded only to those already through the door.

Education in the age of artificial answers

For those still in education, the concern is not just if there’ll be a job waiting – but whether they’re truly learning anything at all. In The Spinoff this morning, Hera Lindsay Bird documents the creeping use of generative AI in New Zealand schools and universities. Some of the quotes from teachers and lecturers are eye-popping. “My main beef with AI is that it made me into a grown adult asshole who had 18-year- old enemies,“ one says. “I wanted to be a teacher not a cop.”

Lindsay Bird writes: “Almost every educator I spoke with, from primary school teachers to those supervising postgraduate dissertations, raised serious concerns,” she writes, “with some teachers estimating that up to 80% of their students relied on ChatGPT to complete assignments.

“I spoke to MA supervisors whose history students theses were riddled with fictitious sources and ‘archival’ Midjourney photographs, and primary and intermediate school teachers, who said students as young as 11 were using it to answer simple personal prompts, such as ‘what did you do in the summer holidays?’ and ‘what was your favourite penguin in the text?’”

Read: Hera Lindsay Bird on the reality of teaching in the world of AI

Keep going!
A long shadow of a person is cast on a sunlit pavement. On the left, an orange vertical banner reads "THE BULLETIN" in white text, with a white arrow above it.
Photo: Getty Images

The BulletinJune 6, 2025

Press secretary case raises questions for police and officials

Contributing writer
A long shadow of a person is cast on a sunlit pavement. On the left, an orange vertical banner reads "THE BULLETIN" in white text, with a white arrow above it.
Photo: Getty Images
newsletter

The disturbing allegations have sparked a review of vetting processes and questions about the adequacy of our laws, writes Catherine McGregor in today’s extract from The Bulletin.

To receive The Bulletin in full each weekday, sign up here.

Allegations prompt questions about police, Beehive processes

The fallout from serious allegations against Michael Forbes, the prime minister’s now-former deputy chief press secretary, continued on Thursday, following a damning Stuff investigation released the night before. Forbes resigned on Wednesday after journalist Paula Penfold put to him allegations that he had taken non-consensual recordings and photos of women, including sex workers, in various settings. His phone reportedly contained covert recordings of sexual encounters, images of women at the gym and supermarket, and photos taken through windows – including one of an unclothed, unconscious or asleep woman. Forbes apologised for his actions and said he had sought professional help last year.

The allegations have sparked widespread concern, with questions raised about the adequacy of the law, the Beehive’s vetting processes, and why no criminal charges were laid.

Legal grey areas highlighted by lack of charges

Despite the disturbing nature of the allegations, police concluded the behaviour did not meet the threshold for prosecution. On the allegation involving an image of an unclothed, unconscious woman, AUT law lecturer Paulette Benton-Greig told Checkpoint it “certainly seems to me that [it] could have been chargeable” under existing laws about intimate visual recordings taken without the subject’s consent.

But as legal academic Cassandra Mudgway explained in the Conversation, republished on The Spinoff, current law does not cover covert audio recordings, nor most intrusive photographs taken in public spaces. This gap in legal protections has prompted calls for reform, including from the sex workers who brought Forbes’ behaviour to public attention. Justice minister Paul Goldsmith said the discussion could be had, but he “wouldn’t underestimate that that’s a big change” to what is settled law.

Vetting processes under review

The investigation into Forbes’ behaviour began in July 2024, months after he had been vetted for his job with minister Louise Upston, and before he was later appointed to the PM’s press team, report The Post’s Thomas Manch and Kelly Dennett (paywalled). Despite having a security clearance that required disclosure of any subsequent police contact, Forbes did not tell his superiors about the investigation. “That didn’t happen,” Luxon said on Thursday, describing himself as “incredibly concerned” about the revelations and acknowledging the “distress” among Beehive staff.

The Department of Internal Affairs has launched an urgent review into the vetting processes and whether inter-agency systems failed. Luxon, while emphasising the importance of police independence, called it a “fair” question to ask how such a case could go unnoticed by senior officials and decision-makers.

Why didn’t police inform the Beehive?

That question has taken on new urgency amid confusion over why the police never alerted the Beehive. Police seized Forbes’ phones and interviewed him, but ultimately decided no offence had been committed, report Manch and Dennett in a separate story (paywalled). Commissioner Richard Chambers said under the “no surprises” convention with ministers, it is up to the commissioner of the day to decide what to escalate. However Andrew Coster, who was commissioner at the time, said he was never informed of the case.

Chambers first learned of the matter only this week when approached by Stuff, at which point he alerted the Department of Prime Minister and Cabinet. It seems that someone senior in the police may have dropped the ball, but Chambers wouldn’t name names, only observing that “it is important that police executive members alert their Commissioner to matters that may need consideration.”