Machines have been accessing our information online for two decades, so why is everyone so het up about it now? Because, the stakes have changed, and while it might not be costing us money, we’re paying for it with something far more valuable, writes FCB Media’s Agata Ziemianska.
In the past few weeks, we’ve seen fresh outrage over AI tools “scraping” personal data. First, it was LinkedIn automatically opting millions into data-sharing for Copilot training. Then, rumours spread that Google’s Gemini would soon be “reading your emails”.
It all sounds invasive, and deeply unsettling. But here’s the uncomfortable truth: none of this is new – and opting out won’t change as much as we might think.
We’ve been here before
If you rewind to the early days of Gmail, Google was astonishingly transparent about what it was doing. Its Terms of Use made it clear that ads would appear next to your emails because Gmail automatically scanned their content. One (now archived) version states plainly that Gmail “serves relevant ads using a completely automated process… No human will read the content of your email.”
In other words, machines have been accessing our personal information for nearly two decades, long before AI assistants, personalised recommendations or targeted ad ecosystems became part of everyday life.
So the question isn’t: “Why is AI suddenly reading my data?” Rather: “Why are we only noticing now?”

The stakes have changed
The difference between 2005 and 2025 is what the technology can do with the data. Back then, automated scanning powered slightly more relevant ads. Today, automated scanning can:
- profile your behaviour
- summarise your emails
- generate content in your tone of voice
- make suggestions
- and, increasingly, make decisions on your behalf
AI has closed the gap between data collected and value extracted. The leap in capability has led to a leap in public emotion.
Data that once merely targeted ads now fuels systems that think, create and predict. That’s why the same practices – now wrapped in AI – trigger a completely different reaction.
The internet has always had a price
One of the most persistent digital myths is that online content is free: news, social networks, search tools, email.
But we’ve always paid for these with something more valuable than dollars: our time, our attention and our personal data. And the rise of AI has supercharged this economy.
Your comments on Reddit become training data for someone else’s multimillion-dollar model. Your reviews train algorithms that recommend products to millions. Your social posts shape the tone and accuracy of AI answers for strangers on the other side of the world. Whether you realise it or not, you are a supplier in the AI value chain.
New Zealand: a high-paywall, low-trust market
It’s tempting to think that if we don’t like the data economy, we can simply choose subscription models instead. After all, New Zealand has one of the highest adoption rates for paid online news globally – AUT estimated 24% of New Zealanders paid for digital news in 2024. But flip that statistic around.
It also means three out of four New Zealanders don’t pay, preferring the “free” version instead. This forces publishers to rely even more heavily on advertising and data-tracking models to survive.
New Zealand is now in a strange position: digital subscriptions are rising, paywalls are tightening, but audience trust is falling and news avoidance is climbing.
Consumers want quality, independence and privacy… but also want content for free. Something has to give.
Human conversation = the most valuable currency
AI systems aren’t powered by glossy marketing websites – they’re powered by messy, honest, human conversation. This is why platforms like Reddit, StackOverflow and X have suddenly become goldmines: their content is created by real people. It’s structured as natural Q&A. It’s raw, specific and emotionally expressive. It’s constantly refreshed.
For AI, this content isn’t “nice to have”. It’s foundational. Human language, in all its messy brilliance, is now the most valuable resource on the open web.
Every interaction: every post, search, review, complaint or email, becomes training material that shapes how AI sees us, advises us and – soon – acts on our behalf. Our contributions influence how information flows, what surfaces in AI search and which voices dominate online conversations.
Opting out of data-sharing settings might protect your personal profile, but it won’t change the underlying reality: AI is hungry and the modern internet is built to feed it.
The conversation we need to have is not: “Should we opt out?” It’s what we want the invisible economy of data, content and AI to look like – and who should profit from it. Because whether we pay with money or data, one thing is certain: nothing on the internet has ever been free. And in the age of AI, the price is only going up.
The post The true price of AI and the internet in 2025 appeared first on stoppress.co.nz.
More Stories
Over $100k initiative powers up Auckland’s indie stages
Acing Brand Experience With Mammut CMO Nic Brandenberger
Outward Bound leans into uncertainty with fearless new platform