Enjoy Your Free Gift
The hidden cost of free services — how consumer data became the currency of the digital economy.
title: "Enjoy Your Free Gift" slug: "enjoy-your-free-gift" description: "The hidden cost of free services — how consumer data became the currency of the digital economy." datePublished: "2012-11-02" dateModified: "2026-03-15" category: "AI & Privacy" tags: ["privacy", "free services", "data economy", "consumer"] tier: 3 originalUrl: "http://www.applieddatalabs.com/content/enjoy-your-free-gift" waybackUrl: "https://web.archive.org/web/20121102214005/http://www.applieddatalabs.com:80/content/enjoy-your-free-gift"
Enjoy Your Free Gift
"If you're not paying for the product, you are the product." That line had been floating around the internet for a few years by the time we wrote this piece in 2012. We used it as a jumping-off point to explain how the free services people loved, Google search, Facebook, Gmail, were funded by an exchange most users didn't fully understand. You got a useful tool. The company got your data. That data was sold to advertisers. Everyone seemed happy, or at least nobody was complaining loudly enough to change anything.
Our argument in 2012 was that this trade was becoming less fair over time. Companies were collecting more data than the services required, holding it indefinitely, and finding new ways to monetize it that users never consented to. The "free gift" was getting more expensive. We had no idea how expensive it would get.
The Deal in 2012
In 2012, the data-for-service exchange was relatively simple. Google read your emails to show you relevant ads (yes, actual humans at Google could theoretically access your Gmail, though the company insisted they didn't). Facebook tracked which pages you liked and who you friended, then used that to target ads. Both companies placed tracking cookies that followed you around the web, reporting back what sites you visited after leaving their platforms.
Most users accepted this without thinking much about it. Free email, free search, free social networking. The alternative was paying for these services, and nobody wanted to go back to paying for email after years of Hotmail and Yahoo Mail. The economics seemed clear: a few targeted ads were a small price for genuinely useful tools.
We pointed out some cracks in this logic. The amount of data being collected was growing fast. Facebook's acquisition of Instagram (for $1 billion in 2012) and its plan to acquire WhatsApp (completed in 2014 for $19 billion) meant the company was assembling a surveillance apparatus that tracked your photos, your messages, and your social connections across multiple apps. Google's integration of data across Search, Gmail, YouTube, Maps, Android, and Chrome created a similarly comprehensive profile.
In 2012, you traded data for free email. In 2026, you pay $14 a month for a streaming service that also sells your viewing habits to advertisers. You've become both the customer and the product.
Paying AND Being the Product
Here's what changed: the old formulation doesn't work anymore because now you're often paying and your data is still being harvested.
Netflix, which charges $15.49-$22.99 per month, uses viewing data to train recommendation algorithms, inform content production decisions, and (since 2022) sell advertising on its ad-supported tier. Amazon Prime costs $139 per year, and Amazon mines your purchase history, browsing behavior, and Alexa voice recordings for its advertising business, which generated $46.9 billion in 2023. Even your car is in on it. GM was caught selling detailed driving data, including location, speed, and braking behavior, to insurance companies through third-party brokers like LexisNexis, and drivers were getting premium increases as a result.
The boldest move came from Meta. In November 2023, facing EU privacy regulations, Meta offered European users a choice: pay around $14 per month for ad-free Facebook and Instagram, or continue using the services for free and consent to behavioral tracking for ad targeting. The EU's data protection authorities challenged this as a false choice, arguing that privacy shouldn't cost $168 a year. But Meta's move made the data-for-service exchange explicit for the first time. Your personal data had a price tag, and Meta was happy to tell you exactly what it was worth.
Your Data Now Trains the AI
The most significant shift since 2012 is that your data isn't just being sold to advertisers anymore. It's training AI models.
Every post you've written on Reddit, every answer you've contributed to Stack Overflow, every photo you've uploaded to Instagram, every email you've sent through Gmail has become potential training data for large language models. OpenAI trained GPT-3 and GPT-4 on huge swaths of internet content, much of it created by users of free platforms who never imagined their words would be used to build commercial AI systems. Google trains Gemini partly on data from its products. Meta openly acknowledged training LLaMA on public Facebook and Instagram posts.
This is the final evolution of the "free gift" concept. In 2012, you got free email and the company got to show you ads. In 2026, your accumulated years of digital activity are being used to build AI systems worth hundreds of billions of dollars, and your compensation remains: free email. The asymmetry has become absurd.
Some people are pushing back. Artists have filed class-action lawsuits against AI companies for training on their work without permission. The Authors Guild sued OpenAI on behalf of thousands of writers. Reddit and Twitter started charging AI companies for data access, though none of that money goes to the users who created the content.
The Enterprise Version of This Problem
Businesses face their own version of the free gift trap. Cloud providers offer generous free tiers and easy onboarding, but your usage data and workflow patterns become part of their competitive intelligence. SaaS tools trained on your company's data get better at serving your competitors. Companies building on third-party AI APIs are sending proprietary business data through systems that may use it to improve models that benefit everyone, including rivals.
Smart organizations are applying data governance principles to these vendor relationships, understanding exactly what data leaves their systems and how it's used. The Operational AI framework treats data flow analysis as a core governance practice, because in 2026, there's no such thing as a free gift in enterprise software either.