Why Data Privacy Is A Lie We Tell Ourselves
…And What We Can Actually Do About It
A few years ago, I was working on a product that required integrating a CAPTCHA solution. You know, those little tests where you prove you’re not a bot by selecting all the traffic lights? Turns out, those aren’t just keeping out spam.
I had lunch with one of the people behind it, and he casually dropped a fact that’s stuck with me ever since:
“In every block of London, there’s a hacker collecting data.”
At first, I thought it was an exaggeration. But the more I dug into it, the more I realized… he wasn’t exactly wrong.
The Great Privacy Illusion
We like to believe that toggling a few settings on our phones, using incognito mode, or rejecting cookies keeps us safe. It doesn’t. Data collection is relentless. Everything from your social media likes to your face in a security camera is being recorded, analysed, and more often than not, sold.
Fact check?
50% of UK businesses reported a cyber-attack in 2024.*
The average cost of a UK data breach? £3.58 million.**
Transport for London had a breach that exposed 5,000+ passengers’ personal data, including home addresses and bank details.***
And that’s just what we know about.
But Here’s the Twist: AI Needs Us
While companies are hoarding data like digital dragons, there’s an ironic problem creeping in: AI is running out of good data.
Human-generated data—real, diverse, high-quality information—is drying up. We are already starting to rely on synthetic data (AI-made data to train AI). But fake data leads to… well, fake intelligence.
That’s where we have leverage.
So How Do We Flip the Script?
Instead of passively giving away our data, we should be taking equity in the systems we’re feeding. AI companies need human contributions…whether that’s text, images, or voice data.
What if we made them pay for it?
Steps to take control of your data (and maybe even profit from it)
Be picky about where you show up. Platforms like TikTok and Meta own anything you post. Stay informed and choose where to post what.
Use data poisoning tools. Programs like Fawkes subtly scramble images so they can’t be used to train AI models effectively.
License your content. If you’re a writer, artist, or photographer, look into data licensing models
Support AI that compensates contributors. Some emerging AI projects are offering revenue shares for data contributions. If they need your work, they should pay for it.
Demand equity. Real protections won’t happen unless people demand them.
We’re past the point of keeping our data completely private. The real question is, do we just accept that we live in a surveillance economy, or do we demand a cut of the profits?
Because right now, the game is rigged. And if AI is going to get smarter off our backs, we might as well start setting the terms.
References:
*https://www.twenty-four.it/services/cyber-security-services/cyber-crime-prevention/cybercrime-statistics-uk
**https://www.smeweb.com/uk-data-breach-costs-soar-to-3-58m
***https://tfl.gov.uk/campaign/cyber-security-incident