Let’s start with three threats: malware, Man-in-the-Middle attacks, and physical access to the device.
Malware is most often used in mass attacks for financial gain. There are many ways to cash in: mining cryptocurrency on infected computers, stealing financial data, encrypting the victim’s disk and demanding money for decryption, and so on. What have become popular lately are clippers, which are malware that change the clipboard. This is how a friend of mine lost money once – a clipper changed his clipboard when he copied a cryptocurrency wallet address.
Although mass attacks using malicious software can cause enormous damage to corporations, individual users have a good shot at protecting themselves from most of them. This book will help you do just that. But bear in mind, it’s “most of them”, not “all of them”.
There is also malware that exploits zero-day vulnerabilities, i.e., vulnerabilities that are not yet known to software or hardware developers and have not been fixed, but are already known and exploited by attackers. If the latter have also happened to find a critical vulnerability in widely used software or hardware, malware that exploits this vulnerability can cost millions of dollars.
Such powerful malware is usually developed by highly competent people from corporations, government agencies, or skilled hacker groups. Infection can take place entirely unnoticed by both the user or antivirus software.
I would say that preventive measures would help here, but that would only be partly true, because sometimes there is little we can do about it. For example, the well-known spyware Pegasus, developed and
sold by the Israeli corporation NSO Group, could (and may still be able to) be installed on iOS and Android via zero-click exploits, i.e., without any user interaction with the device. Back in 2021, which is not so long ago, it was quite possible
1 to get infected with Pegasus on iOS simply by receiving a message with a JBIG2 file in iMessage. Apple has fixed this vulnerability, but Pegasus as a whole is still alive and actively used by you-know-who.
If there is any sound protection against such programs, it would be to simply abandon mobile phones, as it is difficult to achieve the same level of security on a smartphone as on even the most basic Linux PC. Of course, there are many situations in which not having a mobile phone would be challenging. Moreover, there is no guarantee that your PC is safe, even if it has near-perfect software protection measures in place: it should not be forgotten that in addition to software vulnerabilities, there are also hardware vulnerabilities, which are occasionally found on both Intel and AMD processors. In the second chapter, I said that it is not always possible to eliminate or mitigate a threat, and all that remains is to accept the threat.
Bottom line, you can learn how to protect yourself from malware, but just like with anything else, you should never expect a 100% protection. Some people think that the only safe computer is a turned-off computer, but even that idea isn’t totally right, as we’ll see later.
Now a few words about Man-in-the-Middle attacks. The general principle is clear from the name—an attacker takes control of the victim’s communication channel with someone and tracks or modifies the flow of information. Some MITM attacks are now irrelevant thanks to the development of network applications and protocols. At the same time, there is a number of attacks targeting local Wi-Fi and Ethernet networks (the data link layer of the OSI model), the risks of which are often underestimated by ordinary users.
We will discuss such attacks in the next chapter, but for now, I will throw in a bit of intrigue: phishing does not only occur when you click on a fake link that takes you to a fake website, but also when you click on a real link that takes you to a fake website.
Let’s move on to physical access to a device or its memory. At first glance, it seems easy to protect yourself: all you need to do is use reliable encryption technology and a password. This is very important, but encryption is not a magic bullet: when you entered the password, you decrypted the device, and that’s when someone could simply take it away from you. All they need to do is wait for or create the right moment.
First example: you are sitting in a cafe, immersed in some degenerate content on TikTok, oblivious to the world around you. In this situation, it would be easy to sneak up and steal your phone, especially since it is unlocked. Finito. The same thing can happen if you are using a laptop in a public place.
Second example: someone approaches you and asks to borrow your phone for a couple of seconds to make a call or check the navigation app because their phone is dead, and the stranger is late, lost, drunk, etc. You act like a good Samaritan and unlock your phone, but at that very moment, you are restrained by officials and your device is confiscated.
Later, when we move on to deanonymization, you will learn the real story of a criminal whom the cops were able to bring down thanks to a well-organized provocation carried out at the right moment.
In addition to “taking it away after the password is entered,” there is also the option of peeking at the keystrokes as the password is entered. This simple and effective method is great because it can be used not only for physical access, but also for remote access: passwords are not only used to access devices, but also accounts on various online services.
If you often use your devices around strangers, either break this habit or always keep the threat in mind and try to minimize the risks.
To protect against device seizure and snooping, here is some practical advice I can give you right now: first, use your phone at a sufficient distance from other people and security cameras; second, it is better not to use laptops in public places unless absolutely necessary.
Finally, an often overlooked method of gaining physical access to an encrypted device is torture. Legally, of course, the police have no right to torture you, but law and reality are two separate worlds with only a few intersections.
The topic of torture is covered in the fourth section of this book, but the second section also suggests some useful preventive measures against torture, related to computer security.
The next two threats are privacy breach and deanonymization.
To avoid confusion with the concept of privacy breach, we will temporarily exclude deanonymization so that we can look at it separately later. Accordingly, let’s think of privacy breach as a situation where it’s revealed
what someone did, and anonymity breach as a situation where it’s revealed
who did something.
How can our enemies breach our privacy on the Internet? Since in our threat model the main enemy is a cop, we must take into account that the enemy has the ability to file legally justified and not-so-justified requests with a chance of obtaining useful answers.
Do you think there is any reason why IT corporations would not rat you out to the noble police? That’s naive, because they’re already leaking your data, and not just to the cops, but to anyone who wants it.
Most ordinary people use Google, Apple, Microsoft, and Meta products in one way or another. These four giants hold a huge amount of information about literally every one of their users.
Corporations violate privacy primarily for marketing reasons. You may have noticed “coincidences” like this: you google something, and after a while you see ads on the same topic, not on Google itself, but on another platform. Well, these are not coincidences, but data trading and subsequent targeting — identifying a target audience among the general public in order to show targeted ads to them.
That is why, in large-scale marketing, advertisers strive to learn everything about everyone, and for a certain amount of money, they can obtain this knowledge. Trading user data is one of the key sources of income for these corporations.
It would be fine if marketers only tracked Google searches or Instagram subscriptions, but their greed goes much further. A recent example is Meta
selling personal correspondence to Netflix. The link to this story also contains other materials about Meta’s privacy breaches. You can also read about
Microsoft,
Apple, and
Google’s “care” for personal data.
In general, the scale of the problem is already significant, and now consider that corporations would leak no less to the cops than to advertisers. I would like to reassure you that you will learn to cope with this inconvenience, as with the others. For now, let’s continue with a general overview of threats and smoothly transition to deanonymization and fingerprinting.
Deanonymization means figuring out a person’s identity.
Fingerprinting is the process of comparing various online activities and figuring out that they were done by the same person, without necessarily figuring out their identity.
Fingerprinting can be used in targeting or in investigative work as part of the deanonymization process.
I will briefly describe several methods of deanonymization and fingerprinting.
Let’s start with the famous “IP tracking” that everyone has heard of. An IP address is a unique identifier for a device on the network that is necessary for going online. By knowing the IP address, anyone can determine the approximate location of the device without spending any resources, and the cops, with their resources, can go to the provider and obtain a lot of valuable data that will make it much easier to search for a suspect. The threat is serious, but it is relatively easy to protect yourself from it at a basic level. However, there are many caveats, and it is not enough to simply connect with the first VPN you come across. Furthermore, no method of hiding your IP address will guarantee anonymity on its own and is not an ultimate solution.
Your browser can tell a lot about you. So it’s not just about ditching trashy products like Chrome, switching to something like Firefox, and limiting cookies. That will definitely be a step in the right direction, but there’ll still be a bunch of threats you can’t get rid of, and those are called browser fingerprints. You can learn more about this with
browserleaks or
amiunique, where you can also see how these fingerprints help to identify you personally. Browser fingerprints are primarily used for targeting and blocking scammers, but there is absolutely nothing stopping them from being used in operational and investigative work.
There are also hardware-level identifiers, such as MAC addresses and various serial numbers like IMEI. Under normal circumstances, these identifiers cannot be changed, so you need to be especially careful when trying to hide them.
Now let’s move on to the reason why all the deanonymization methods described above are not the worst. The reason is the perpetrator. Often, the simplest and at the same time most effective approach to deanonymization is to use the information that the perpetrator reveals about themselves through their own words or actions. The main enemy of your anonymity, my friend, is you yourself. Let’s turn to historical experience.
Our patient is Ross Ulbricht, founder of Silk Road, the then largest darknet marketplace selling drugs and other illegal goods. He was a thoroughly repulsive individual, but not stupid, and he was quite good at keeping his cover up. So what led to us now knowing him by his real name, Ross Ulbricht, rather than his pseudonym, Dread Pirate Roberts?
Below, I will briefly summarize the most important points. For those who want to read the whole story in detail, here is a link to the
article by the cop who caught Ulbricht.
Using the account of an arrested girl with moderator rights and skillfully pretending to be her, the cops managed to establish that Silk Road’s servers were located in Iceland. This was a major breakthrough, but it did not lead directly to the ringleader. All they managed to find out was that the administrator accessed the server from an internet café in San Francisco.
While communicating with the administrator, nicknamed Dread Pirate Roberts, the cops noticed that his time zone was displayed in the chat—Pacific Time Zone, which is the time zone used in San Francisco.
At the same time, the officers conducted a thorough Google search and found a forum, a gold mine, I would say, where Silk Road was mentioned back in January 2011. Ross Ulbricht’s email address was found in posts mentioning Silk Road, and on the same forum, Ulbricht asked questions about Bitcoin and Tor, the very technologies on which his market was based and which, at the time, were not nearly as popular as they are now.
When Ulbricht was checked against police databases, it turned out that the police had previously dealt with Ulbricht because of forged documents. When asked about the source of the documents, Ulbricht stated that he had purchased them on Silk Road.
In short, Dread Pirate Roberts was deanonymized, and the rest was a piece of cake.
The FBI considered simply storming Ulbricht’s home with special forces, but this was risky because the huge metal gate would allow Ulbricht to react quickly and remove all evidence.
They decided that a more cunning approach was needed, and they found one. Ulbricht left his house with his laptop, and his journey was tracked. He went to the library to use the public Wi-Fi there. Ulbricht was no fool and used encryption, so just taking his laptop wasn’t an option. However, encryption doesn’t help if you’ve already opened and, therefore, decrypted something.
A cunning cop, who at that time had an admin account on Silk Road, wrote to Ulbricht and provoked him into opening everything that was needed. Then, to be on the safe side, a couple of other agents pretended to start a fight in the library to distract the dealer from his laptop. Ulbricht fell for it, and before he knew it, one agent jumped out and grabbed the laptop, while another grabbed Ulbricht himself.
By that point, there was already ample evidence of his guilt. Another nail in Ulbricht’s coffin was that he promoted libertarian ideas from both his real accounts and under the guise of Dread Pirate Roberts. Most likely, he sincerely believed in the moral justification of his actions. Ulbricht once wrote on his LinkedIn page:
“I am creating an economic simulation to give people the opportunity to experience for themselves what it is like to live in a world without the systematic use of force.” For my part, I would add that, in addition to drugs, another thing that was distributed in this “world without the systematic use of force” was child pornography.
Silk Road was succeeded by another darknet marketplace, AlphaBay. I won’t retell this story here, but I recommend
reading about it yourself in your free time. AlphaBay was much larger than Silk Road, but its creator made much more serious mistakes. For example, for some time he put his personal email address in welcome messages to new AlphaBay members.
Let’s move on to the conclusions.
Ross Ulbricht made one seemingly minor mistake — long ago, he left several messages on a forum from an account linked to a non-anonymous email address, and that became the beginning of the end.
To achieve anonymity on the Internet, you must separate your virtual identity from your real one, as well as separate your virtual identities from each other if you have more than one. To do this, you need to be aware of a number of vulnerabilities, not only technical, but also social or socio-technical in nature.
1. Account details. This includes your email address, nickname, and password. For the first two, it’s obvious: don’t use the same email addresses and nicknames, because they make it easy to find you.
Passwords are also straightforward. First, when creating passwords, a huge number of people use their personal data, which, incidentally, also makes the password itself weak. Second, using the same password for multiple accounts carries dual risks. If one account’s password is leaked, the other accounts can be accessed immediately. Additionally, the leaked passwords can be compared, revealing that different accounts belong to the same person.
2. Money. It is clear what the problem is with banking services: anonymity and often even confidentiality are simply impossible there (we are not considering the competent use of front men, as this is currently available to very few people, and every homeless person already holds dozens of payment cards).
On the other hand, cryptocurrencies give many people a false sense of invulnerability. In fact, cryptocurrencies can only be independent of external regulators because they use a special technology that requires ALL transfer records to be stored at ALL network nodes. In a sense, the balance of any crypto wallet is not a balance in the usual sense, but a history of all transfers to and from it, and this history is publicly available to anyone who wants to see it. One might think: no big deal, you can’t tell who exactly is making the transfers anyway. Unfortunately, it is possible to find out.
If you use cryptocurrency, you most likely buy or sell it somewhere for fiat money. This is where you can be caught: most converters and stock exchanges operate within the legal framework. Moreover, according to our information from one law enforcement officer, converters and stock exchanges may periodically leak to the police everything they know about all their customers, even without specific requests for individual suspects.
It is possible that you are a more advanced user and use mixers, but this does not solve the problem either: a mixer can also leak your information, and I am convinced that there are even more honeypot mixers than normal ones. In any case, you have no way of making sure that a mixer is reliable.
But don’t be discouraged: it is possible to anonymize crypto, and in one of the following chapters, I will explain how to do it.
3. Social circle and field of activity/interests. If your virtual persona has common contacts with your real persona, then for obvious reasons this can also be a factor in de-anonymization. Of course, this is not a call to automatically break ties with all your friends who, for whatever reason, already know the details of your real identity. Just give some thought to how you will work going forward. Perhaps you know each other within your circle, but in the future you will be able to remain anonymous to everyone else and avoid creating new intersections between your real and virtual personas. And don't forget that it's not just personal data itself that needs to be protected, but also the fact that someone knows someone else's personal data.
When it comes to fields of activity and interests, the importance of such information is usually higher than it seems. Here's an example: let's imagine that in the fall of 2023, a certain perpetrator blurted out that he was a first-year university student. In 2023, 267,000 candidates
applied for undergraduate programs at Ukrainian universities. This may seem like a lot, but in reality, it greatly narrows the search. Now let's assume that our offender mentioned his major, “Computer Science.” There were only 30,000 applications for this major that year (and this is a popular major in Ukraine; for unpopular majors such as chemistry and physics, the number would be significantly lower). Let's add that the offender mentioned living in Kyiv. There are not many higher education institutions in Kyiv, and here we can immediately exclude many universities that do not offer the specialty we need. Finally, if the perpetrator also mentioned his university, the investigator only has to go through a few hundred people, which is very few.
Note: in this example, we only used data about the field of activity and city of residence. If we add everything else to this, the pool of suspects can quickly narrow down to a single person. After all that has been said, I must reveal a secret: the above was not an imaginary case, but my real experience of deanonymization, even though I am just an ordinary guy from Podunk with no police resources or connections: only a console, a monitor, a mouse with a keyboard, and Internet access. I didn't even buy any restricted databases; I found everything from public sources.
By the way, I mentioned my hometown of Podunk on purpose. I lied—it's not my hometown. Remember that when you use a virtual persona, you can not only conceal your real personal information, but also replace it with misinformation. You need to do this carefully, not too intrusively, as if you let it slip. Let your ill-wishers think that you made a mistake and that they were lucky to notice it.
4. Language use — another point worth emphasizing, though often overlooked. Each of us exhibits a unique combination of features, such as overall literacy level, punctuation, use or avoidance of emojis, frequently repeated phrases, capitalization, and more.
None of these things are particularly significant on their own, but together they can be quite important. It is not a given that law enforcement specialists will conduct a linguistic analysis of your writing, nor is it a given that such an analysis will yield any useful information, but the possibility exists and is worth keeping in mind. Sometimes it makes sense to deliberately change your patterns. For example, if your grammar is flawless, your style is elegant, and you never use emojis, start writing with errors, awkward phrasing, and lots of emojis to create an impression of mild stupidity. Or maybe not so mild.
5. Voice. I don’t know about other countries, but in Ukraine, it would be difficult to identify a person by their voice, unless, of course, their voice is publicly known as the voice of Mr. So-and-so. Another thing is that after the criminal has been caught, the voice can serve as valuable evidence if incriminating records exist. Fortunately, we live in the age of neural networks, so you can suggest to your lawyer that the evidence be proven invalid, because now it is quite possible to vocalize any text in any voice, and it will sound natural.
Hypothetically, such a trick could work if your case is dealt with strictly within the legal framework (in Ukraine, that alone would be a stroke of luck), the other evidence is weak, and your lawyer is strong.
However, it is better not to let it come to that and not to test this idea on yourself. When you need to say something anonymously but publicly (for example, for a YouTube podcast), those same neural networks are very good at replacing the voice on the recording. You can even make yourself sound like Mr. Krabs or Geralt of Rivia.
Real-time voice modification is more challenging, as software tends to produce poor-quality output. Much depends on your individual characteristics: the same software might work well for one person but not for another. Experiment to find what works for you, preferably use a state-of-the-art AI voice changer. Unfortunately, I cannot share my experience here, as it has been a long time since I last used voice changers.
That’s all for now on confidentiality and anonymity. We will discuss specific types of attacks and methods of protection in more detail in the corresponding chapters, but for now, let’s transition to the last threat in this chapter with another story.
One autumn day, Sasha came back home from classes to his rented apartment on the first floor of a Khrushchev-era building. The young man crossed the doorstep, took off his Gucci sneakers, walked into the room with old, peeling wallpaper, and turned on his new gaming PC. In the lower right corner of the desktop, Sasha noticed that the Internet had disappeared.
When Sasha called his provider, they replied briefly: a technician would come today and fix the problem.
Half an hour later, the technician arrived and immediately made a positive impression on Sasha. While the specialist was busy with the problem, the young man went out onto the balcony: he didn’t want to stand over the nice man’s shoulder. The technician fixed everything, said that the Internet would be back soon, received his well-deserved payment, and took his leave.
An hour later, Sasha heard another knock on the door and saw another technician. The student recapped the recent events, and the second technician left, cursing under his breath, while Sasha returned to his room and continued chatting with his friends on Telegram.
You may have already guessed what really happened on that autumn day. Now I will tell the same story, but from the perspective of other characters.
Alexander Denisovich Petrenko had been under investigation by the authorities for some time. Young investigator Andreev felt that Alexander was not only an interesting person, but also someone who could help him advance his career. Andreev’s confidence grew when Petrenko, who had previously been poor, started boasting about expensive purchases on his social media. However, although the prosecutor shared the investigator’s suspicions, he said outright that there was not enough evidence yet.
Andreev came to the conclusion that since surveillance and wiretapping had not been very useful, and the investigating judge was not issuing any new orders, it was necessary to use tactical cunning. He didn’t have to think long — there were long-established methods for speeding up the case.
The investigator knew that Petrenko, for some reason, hardly ever missed classes — apparently, he didn’t want to upset his parents. Accordingly, he knew exactly when Petrenko would not be at home. At that very moment, field agent Ivasiuk dressed up as a simple working man, took a suitcase and approached the grannies sitting on a bench, who, after a short conversation, kindly unlocked the entrance for him. All that remained was to find the right internet cable and carefully cut it, which was no problem for Ivasiuk.
When Petrenko came home, he called his mobile provider. The moment could not be missed, and the technician Ivasiuk quickly came to the rescue. He was polite and friendly, creating the image of a reliable professional, thanks to which he hit the jackpot when Petrenko went out onto the balcony: now he had the opportunity to choose a good place to install the bug. After that, the technician told Petrenko that everything would be working soon and left the apartment. Ivasiuk kept his promise and repaired the cable.
They didn’t have to wait long for Petrenko to say something interesting: when he was at home, he conducted business via calls on a popular messaging app. A couple of weeks later, the investigation managed to locate the drug warehouse.
You can imagine what happened next, so I’ll end my story here.
“But the evidence was obtained illegally!” some might say, and they would be right. Anything obtained through illegal wiretapping cannot be used in court. However, such evidence can be used during the investigation, and for the trial, investigators can come up with a completely different, clean story. A simple and effective way to do this is to use well-trained “witnesses.”
So, the most dangerous threat is social engineering, i.e., high-level psychological manipulation aimed at getting the victim to blurt out or do something they shouldn’t.
Most people are vulnerable to social engineering. And the previous sentence was a test: if you just thought that you are not one of the majority, then you are very vulnerable. In fact, everyone is susceptible to social engineering attacks to some extent, and especially those who underestimate them.
Protecting yourself from social engineering is difficult because it requires more than just knowledge and skills: it also requires specific habits.
The fact is that people are trusting by nature and learn not to trust as they grow older. The ability to be reasonably distrustful may not develop until one’s hair turns gray. Why? Some people lack experience, others do not learn from their experience, and still others learn solely from their experience.
To all three categories, I recommend reading The Art of Deception by Kevin Mitnick, a well-known hacker. The book is good because it provides a large number of examples of social engineering, from which one can draw many valuable conclusions. Incidentally, Mitnick argued against the popular saying that a secure computer is one that’s turned off. He said that’s not true either because an ill-wisher could “talk someone into going into the office and turning that computer on”.
As you read this, keep in mind that there are plenty of cops out there who, while not necessarily intelligent, are cunning and highly skilled in social engineering, which is a very important professional skill.
So, what is reasonable distrust? In my view, at its core, it is essentially the same as reasonable trust. Reasonable trust and distrust mean that trust is not your default state, applied to everyone around you unless they give you reason not to trust them. On the contrary, here the whitelist principle applies: you do not trust anyone except those people whom you have consciously decided to trust, with a clear understanding of the reasons behind your decision. This doesn’t mean you see every stranger as an agent of the Sacred Brotherhood of Ukraine; it means you simply know how to clearly distinguish between friends and strangers and don’t grant the latter the same access or information as you do to friends.
The principle of reasonable distrust also applies to verifying the identity of the person you’re communicating with online: you shouldn’t assume they are who they claim to be until you have proof. For example, if someone messages you from an unfamiliar account using a friend’s username, ask them to record a voice message or say something that only you and your friend know. These are the simplest ways to verify the sender’s identity without deanonymization.
To conclude, I’ll say that the principle of reasonable distrust should be applied not only to people but also to software; that’s what we’ll cover in the next chapter.