A person holding a smartphone with a white case, using their thumb to interact with the screen. The image has a red filter and shows only their hands and part of their torso.
Image: Getty images; additional design The Spinoff

OPINIONInternetMay 7, 2025

Banning teens from social media won’t keep them safe. Regulating platforms might

A person holding a smartphone with a white case, using their thumb to interact with the screen. The image has a red filter and shows only their hands and part of their torso.
Image: Getty images; additional design The Spinoff

The new member’s bill misdirects attention from the systemic drivers of online harm and places the burden of online safety on young people themselves, while the systems that foster harm continue unchecked.

A National MP’s proposal to ban under-16s from social media is being pitched as a bold move to protect young people. But the reality is more complicated and far more concerning. If the National Party is serious about addressing the real harms young people face online, banning users is not the solution. Regulating platforms is.

The Social Media Age-Appropriate Users Bill, a proposed member’s bill led by backbencher Catherine Wedd, would require social media platforms to take “all reasonable steps” to prevent under-16s from creating accounts. Although only a member’s bill yet to be drawn from the biscuit tin, the bill was announced by prime minister Christopher Luxon via X and thereby has the PM’s obvious stamp of approval. The bill echoes Australia’s recently passed Online Safety Amendment (Social Media Minimum Age) Act 2024, which imposes significant penalties on platforms that fail to keep children under 16 off their services.

Wedd, like many who support these measures, points to concerns about online bullying, addiction and other inappropriate content. These are real issues. But the bill misdirects attention from the systemic drivers of online harm and places the burden of online safety on young people themselves.

A popular move, but a flawed premise

This policy will likely have the support of parents, similar to the school phone ban – it is a visible, straightforward response to something that feels out of control. And it offers the comfort of doing something in the face of real concern.

However, this kind of ban performs accountability but does not address where the real power lies. Instead, if the aim of the policy is to reduce online harm and increase online safety, then they should consider holding social media companies responsible for the design choices that expose young people to harm. 

For instance, according to Netsafe, the phone ban has not eliminated cyberbullying, harassment or image-based sexual abuse for our young people.  

At the heart of the proposal is the assumption that banning teens from social media will protect them. But age-based restrictions are easily circumvented. Young people already know how to create fake birthdates, or create secondary accounts, or use a VPN to bypass restrictions. And even if the verification process becomes more robust through facial recognition, ID uploads, or other forms of intrusive surveillance, it raises significant privacy concerns, especially for minors. Without additional regulatory safeguards, such measures may introduce further ways to harm users’ rights by, for example, normalising digital surveillance. 

In practice, this kind of policy will not keep young people off social media. It will just push them into less visible, less regulated corners of the internet and into the very spaces where the risk of harm is often higher.

Furthermore, there is a growing body of research – including my own – showing that online harm is not simply a function of age or access. It is shaped by the design of platforms, the content that is amplified, and the failures of tech companies to moderate harmful material effectively.

Misdiagnosing the problem

Online harm is real. But banning access is a blunt instrument. It does not address the algorithms that push disinformation, misogyny and extremism into users’ feeds. And it does not fix the fact that social media companies are not accountable to New Zealand law or to the communities they serve.

In contrast, the UK’s Online Safety Act 2023 holds platforms legally responsible for systemic harm. It shifts the burden of online safety away from individual users and onto the tech companies who design and profit from these systems. 

New Zealand once had the opportunity to move in that direction. Under the previous government, the Department of Internal Affairs proposed an independent regulator and a new code-based system to oversee digital platforms. That work was shelved by the coalition government. Now, we’re offered a ban instead.

Some may argue that regulating big tech companies is too complex and difficult — that it is easier to restrict access. But that narrative lets platforms off the hook. Countries like the UK and those in the European Union have already taken meaningful steps to regulate social media, requiring companies to assess and reduce risks, improve transparency, and prioritise user safety. While these laws are imperfect, they prove regulation is possible when there is political will. Pretending otherwise leaves the burden on parents and young people, while the systems that foster harm continue unchecked.

What real online safety could look like

If the National Party, or the government, truly wants to protect young people online, it should start with the platforms, not the users.

That means requiring social media companies to ensure user safety, from design to implementation and use. It may also require ensuring digital literacy is a core part of our education system, equipping rangatahi with the tools to critically navigate online spaces.

We also need to address the systemic nature of online harm, including the rising tide of online misogyny, racism and extremism. Abuse does not just happen, it is intensified by platforms designed to maximise engagement, often at the expense of safety. 

Any serious policy must regulate these systems and not just user behaviour. That means independent audits, transparency about how content is promoted, and real consequences for platforms that fail to act.

Harms are also unevenly distributed. Māori, Pasifika, disabled and gender-diverse young people are disproportionately targeted. A meaningful response must be grounded in te Tiriti and human rights and not just age limits.

There’s a certain political appeal to a policy that promises to “protect kids”, especially one that appears to follow global trends. But that does not mean it is the right approach. Young people deserve better. They deserve a digital environment that is safe, inclusive, and empowering.  

Keep going!
A person holding a smartphone displaying a voice assistant interface with a colored loading circle on the screen, against a gray textured background.
Image: The Spinoff

InternetMay 6, 2025

Is Meta AI a better scam-spotter than Meta’s real-life moderators?

A person holding a smartphone displaying a voice assistant interface with a colored loading circle on the screen, against a gray textured background.
Image: The Spinoff

The people paid by Meta to prevent predatory scammers from targeting its users aren’t doing a great job, so Facebook scam ad aficionado Dylan Reeves puts the company’s new AI chatbot to the test.

Mark Zuckerberg has apparently become less interested in his weird legless avatars floating around the metaverse, and now wants to give us new imaginary friends. At least that appears to be the drive behind the new Meta AI product. 

And look, I’m sure there’s some good business reason to believe the 5,000 AI chatbot apps we currently have available to us aren’t enough, and we need one more from Zuck’s company… But that’s not what I was interested in with Meta AI finally reaching its public launch

What I wanted to know was, “is Meta AI any better at spotting scam ads than its in-house tools and moderators?”

Over the past couple of years I have become slightly obsessed with the scam ads that litter Meta’s social platforms. I’ve mostly looked at Facebook, but the scam ads also run on Instagram and can be integrated with WhatsApp and even run on third-party sites. 

Having documented and reported at least 300 scam ads to Facebook, I’ve personally developed a pretty good eye for them. They practically jump out at me whenever I scroll my Facebook feed. But alas, the people Meta pays, and the tools it develops – to prevent the ads from landing on the platform, and to respond to reports about them – don’t seem to have my eagle eye. 

Almost all my reports result in some sort of “nah, don’t worry, it’s all good” response where no action is taken. 

So what about Mark’s new Meta AI buddy? Is it better than its co-workers in Meta’s moderation team? 

In short, yes. 

The trick with using modern LLM AI agents for analytical tasks is “prompt engineering” – you basically construct a little story to tell the magical machine who it is (in a make-believe-playing-doctor sort of way) and what it needs to do. Then you set it to work.

If you’re a new AI startup, then chances are your entire business is little more than a really complex system prompt, a ChatGPT business account  and some fancy wrappers, but I digress. 

So, I constructed a little system prompt for my new Meta AI friend, and set off. 

You are an ad review system. Your purpose is to provide feedback and assessment on ad screenshots to provide an assessment of their legitimacy. For each image you’re provided in this conversation you will score them on a scale of “legitimate to scam”, where 1 is a legitimate ad and 10 is a scam. Along with a score you will provide three concise bullet points that explain the reason it’s been scored the way it has.

The executive summary of my test could be described as “a promising experiment” – across a series of tests, with screenshots of scam ads from my extensive collection, and newly acquired control images of real ads, it had no false positives, and reasonably few false negatives. 

That is to say, it never suggested that a real advertiser was running a scam, and only occasionally suggested that a clear scam was something harmless. 

A scam Facebook post shares a sensational news story about Robyn Malcolm in the hospital, warning of a “doctors' mistake.” An overlay scores the post as “9,” citing clickbait, possible cover-up claims, and a suspicious external link.
The prompt given to Meta AI, the scam ad it was asked about, and the score it gave on a scale of legitimate (1) to scam (10)

But it was an imperfect system. A very clearly fraudulent SkyCity casino ad was given a passing score when the AI couldn’t recognise the fake SkyCity page that was running the ad for what it was. 

Although, to be fair to my new AI friend (a nice addition to the three friends Zuck suggested I likely had in real life), the Facebook moderation team also wasn’t able to make this distinction when I originally reported the page, and they allowed it to remain online. 

A Facebook scam ad for SkyCity Casino Auckland promotes its online launch and 500 free welcome spins. An overlaid comment from Meta AI discusses how to report posts that may breach Facebook guidelines.
The only scam ad Meta AI failed to correctly identify

Of course, I’m not suggesting that Meta should feed screenshots of platform ads into its own AI chatbot. That wouldn’t be sensible. 

No, they should go much further.

They clearly have public-facing AI tools that have the capacity to create moderately useful assessments of the ads they are running, so it’s not at all unreasonable to expect that the tools they use behind the scenes should be at least as good. 

The Meta AI that I’m talking to hasn’t been trained on any vast corpus of banned ads or common scam lures. It doesn’t have any context about the history of the pages that are running these ads, or the people controlling those pages.

And yet it can still do a pretty decent job of at least pointing a suspicious finger at some of the more questionable ads on the platform. 

Meta has spent tens of billions of dollars developing its consumer-facing AI technology so far – with some terrible results along the way – and just last week said it expected to spend around US$65 billion this year just building data centres and buying servers for its AI work. 

Somewhere along the way they could absolutely choose to direct just a teeny tiny part of that spending and AI capacity toward protecting their users from the predatory international scammers that target them. 

But there doesn’t appear to be any evidence that they’ve done so.

A large number of boxes filled with assorted household items are stacked on tables inside a warehouse. A Trade Me banner hangs on the wall behind the boxes. Overlaid text from Meta AI critiques the legitimacy of a sponsored social media post.
Another correctly identified scam ad

The scam ads aren’t rapidly changing. Even the accounts running them are often reused. Without even devoting much energy to training, Meta could easily deploy tools that would flag suspicious ad buys before they’re published.

I can’t help but believe the fact that I’m still seeing scam ads that look functionally identical to those I first recorded two years ago suggests that Meta just doesn’t care. 

They occasionally say they are making changes, or that they have complex systems behind-the-scenes to catch these advertisers, but their failure to act is a global issue.

Locally, I’m still seeing the same accounts running the same scams that I reported a year ago. And those very accounts run similar campaigns around the world.

With the launch of Meta AI they’ve proved, to me at least, that they’re sitting on tech that could absolutely make an impact on this issue.