At what point strengthening moderation results in loosening data privacy

Kindness with a catch

Starting November 17th, YouTube is introducing new rules to protect underage users. In short, shielding minors from content involving violence, gambling, so-called social casinos, and digital items of material value is taking a stricter approach. The initiative is generally useful: children need to be protected from the uncontrolled flow of information until they learn to think for themselves. However, there are hidden nuances. You will need to send a photo of your documents to confirm your age, should there be any doubts about the accuracy of the information provided in your account. So, what's the big deal? You need to confirm that you are of legal age, so you confirm it. It’s not the first serves that requires you to send photos of your documents or even a photo of yourself holding it — but there is more than meets the eye. Let’s take a closer look.

Side effect

Of course, children need to be protected from inappropriate content. However, by presenting your documents even to YouTube to confirm your age, your previously anonymous account can now be linked to your real identity. Until now, you could use any name you wanted when creating an account, and no one would check it. But if you provide identification, your account is guaranteed to be associated with you.

This link immediately increases the value of your information for any service. Not only will your data be there, but also all the information about your use of the service. Your preferences, schedule, interests, social patterns, character, all the comments you have left below videos — everything you have done on YouTube in your entire life.

If back in the day it was just some unattributed account, now you've exposed yourself. It’s even possible to look at your activity on other socials, compare your interests with the videos you've been watching, and then cross-reference all of that with your friends.

It is not all that harmless as it seems. Naturally, this data is primarily gathered for marketing purposes, to better promote goods and services. Surely you’ll agree that it is much easier to advertise your products, when you know what your target audience prefers.

Companies have always gathered information about their customers, which is quite normal. This information is typically outlined in the terms and conditions — the agreement where you have to tick a box confirming that you have read and accepted the terms, although almost nobody actually reads them. Sometimes, these agreements contain interesting clauses, such as agreeing that your activity information can be used by the company in good faith and in accordance with the law. However, this refers to the laws of the country where the company is registered, not your own, and you likely have little understanding of how they might use this information. However, these are relatively minor issues. Since the company possesses the data, it will use it in some way or another. The real problem lies elsewhere.

A new player

Until now, people have primarily worked with data arrays, but neural networks have now taken the lead. They thrive on structured and annotated databases. By verifying your identity, you essentially confirm the accuracy of all your data that can be used to train neural networks. How the owners will use a neural network trained on the habits and preferences of hundreds of millions of people is anyone’s guess. On a broad scale, artificial intelligence (AI) works remarkably well with statistics, and there will be an enormous quantity of data about you. Don’t assume you’re so unique that entire corporations will target you personally to persuade you to buy a new product. The reality is far more complex and intriguing than that.

People are far more alike than they think. Everyone can be divided into approximate groups, and work can be done with these groups accordingly.

This isn’t necessarily about advertising. Any idea can be proposed in this way. It will appear completely appropriate and natural to you, presented exactly in the way you’re accustomed to, tailored just as you prefer, and expressed in language you find believable. This is about conveying any narratives in the most acceptable form and manner for a particular audience. While this has been done before, the difference between a group of sociologists and a trained neural network is like comparing a digging stick to a walking excavator. When you verify your real identity with documents on foreign platforms, you are actually helping to build databases used to train such tools.

How to live from now on?

Time brings its changes. Things will never be the same as they once were. Information about you can be both valuable and dangerous. On its own, it might not hold much significance, but when combined with data about others, it can become a powerful weapon. I’m not saying you should stop using the services you’re accustomed to. Just remember that every action you take can now be used to influence you in various ways. If one day you encounter an idea from multiple online sources that seems absolutely right, perfect, ideal and obvious, take a moment to pause and reflect: could it be that, based on analysing your habits and preferences, you were simply shown a thought that aligns perfectly with what you like? When you view it this way, the illusion fades instantly, and you clearly see your own preferences and impressions wrapped around what is being presented to you.

After all, humans are rational creatures, whereas a neural network is simply a very complex formula, nothing more.

Everything mentioned above is not science fiction or some far-off future — it already exists and is operational. If you understand what’s going on, these influences will have much less power over you. Yes, we can no longer completely avoid social networks or stay entirely anonymous on them; our data will be collected and used against us regardless. This is true not only for YouTube — almost all platforms collect user data and use it in some way.

Try to make their job as difficult as possible. If you apply your own intelligence, artificial intelligence won’t be able to manipulate you.

TO THE POINT

Under Belarusian law, personal data cannot be transferred to all countries. Moreover, there are several requirements that must be fulfilled for such data transfers to be lawful. Although it’s not possible to cover every detail here, be aware that your actions might breach the Law On Personal Data Protection. You may not even know the country where the server storing photos of your documents is located. In some cases, authorisation is necessary for cross-border data transfers. While this does not apply in most instances, it is important to keep these considerations in mind.

By Yuri Terekh