Article List

What is the Dark Web?

Many people have heard of the dark web. Unlike something like TCP/IP, which everyone uses on a daily basis but many have never heard of, the dark web is unique in that it has much more people who have heard of it than those who have actually used it. This is in part due to its perception as illegal and scary, but also because it ultimately functions differently to the usual web, so unless you actually want to use it, it’s not something you’ll stumble across by accident. In this article, I would like to explain what the dark web is, how it differs from the deep web and the surface web, and why you (yes, you!) might want to use it.

To begin with, let’s answer a question – what is the dark web? A popular picture used to describe it is that of an iceberg, where the top, the part above the surface, is the surface web, below that comes the deep web, which makes up most of the iceberg, and at the bottom end there’s the dark web, so deep that light doesn’t reach it. This image is often used in conjunction with phrases like “90% of the web is the stuff we cannot see” and “The scary, hidden part of the web is way larger than the internet we normally use”. Both of them are true, but only partially.

The proper definition of the deep web is that it’s content not indexed by search engines. If you search for “Mastodon” on Google, you’ll find the main site with many instances listed. It’s indexed, available to find with a search engine, therefore making it part of the surface web. Once you pick an instance and register or log in, however, you enter the deep web. No matter what you search for, you won’t find my (or anyone else’s) Mastodon account settings. The same is true for Facebook, Amazon, YouTube, and pretty much any site that lets you log in. Going further, consider that YouTube’s homepage, while technically part of the surface web, also has an aspect of the deep web. Google, or any other search engine, cannot index the videos that appear on it, because they change with time, location, whether you’re logged in or not, and so on. Every single unlisted or private video is on the deep web, too. Then there’s websites which can be indexed, but just haven’t been, yet. An example of that is my site; right now, looking up “cirno digital” will give it to you as the top result, but when I initially made it, the crawlers hadn’t yet found it, so no search engine had it listed. Freshly uploaded videos and such also fall into this category. Try uploading a video on YouTube and immediately search its name on Google – you almost certainly won’t find it. While sites not indexed due to time are not very common, if you just think about how much content you interact with every day that’s personalised and/or requires authentication to access, it’s easy to see how “most of the web” would be unavailable to indexes.

Then, there’s the nifty part – the dark web. What’s makes it dark is that it is not really a part of the web. The world wide web, that is. Like the deep web, it cannot be indexed by search engines, but if you know the password or the link, you can reach the deep web through a conventional web browser. What’s special about the dark web is that, even if you know the address, you cannot reach it without special tools, plug-ins or custom software. Because of that, it’s not really a lower section of the web, as much as it is an alternative to the web. It’s not something you stumble across; you won’t get there unless you want to get there. While the dark web is often associated with criminals, this doesn’t have to be the case. Gopher, for example, is an alternative network that started in the early 1990s. It has the same goal as the world wide web; it was just designed differently, and therefore you cannot access a gopher server through your web browser (designed for WWW) without modifying it in some way. It’s also not suited for criminal activities, because it lacks any sort of encryption or protection. Most protocols used nowadays do offer security and anonymity, and they are used by criminals, but it’s a by-product of the services they offer, not the intended, desired use for them. With that said, let’s go over some examples of dark web services that you can use right now.

By far the most popular kind of dark web are the Tor hidden services. Tor stands for The Onion Router, which itself is the name of a special protocol that’s used. When you connect to a server, you’re in a direct connection to it (ignoring lots of complexity here, but bear with it); you know who the server is, and the server knows who you are. What Tor does is it routes your traffic through a number of nodes before reaching the server, and the system makes it so that you’re aware of who all the nodes are, but each node, including the server, only knows of the nodes next to it. Let’s say that there’s me, three nodes A, B and C, and a server that I’m connecting to. I know who A, B, C and the server are; A knows only who I am and who C is; B knows only A and C; C knows only B and the server, and the server knows only who C is. Furthermore, the whole thing is encrypted several times, so a message leaving your end only becomes readable once it reaches the server (and in some cases once it leaves the last node). All in all, you can probably see how this setup is very secure and largely anonymous. Hidden services are a variant of this routing system, which sort of turns both you and the server into nodes, with 6 extra nodes between you, but you only know the 3 nodes on your side, and the server only knows the three nodes on its side. There is a way to connect to the Tor network by installing some software and configuring your browser, but the preferred way is to use the dedicated Tor Browser, developed and supported by the Tor Project, a non-profit foundation. Hidden services all come with a .onion domain, and are relatively easy to set up. The entirety of the privacy and anonymity associated with this kind of the dark web stems from the onion routing method used. It is theoretically possible to break it, but it’s not easy, and other methods have proven more useful, such as users accidentally revealing some personally identifiable information, or being the only Tor user on a given network. This makes it worth pointing out that Tor works best when lots of people are using it. If there’s only one person who visits a certain site through Tor, even if the server doesn’t know who that person is, their presence is still easy to detect, because every time a Tor node connects (those are publicly known), it’s got to be that person, which opens them to all sorts of other de-anonymisation attacks.

Another network we will take a look at is called Freenet. You can use it from your regular browser, but you first have to set up a node on your computer, which you do by installing the software from the project’s page. Freenet’s uniqueness stems from the fact that is a decentralised, peer-to-peer network. While the conventional internet and Tor operate with servers hosting and sending their data to others, Freenet consists entirely of nodes, where each holds little bits of information on the more popular sites. And so, when you want to open a site, you announce that interest to the nodes you’re connected to, and they scrounge up bits of the site and send it to you, until you have enough pieces to display it properly. This system should seem familiar to anyone who’s ever used torrents to download things. This has both its benefits and its drawbacks. Arguably the largest benefit is that there does not exist a single server to bring down. At best, you can bring down a node or two, but the website will still exist in little pieces, spread out across thousands of other nodes. This makes the network extremely resistant to censorship, making it ideal for activists, whistleblowers and any other people who might wish to say things that offend some Big Brother. The biggest weakness is that the health of a website or a file (known as an “insert”) depends on its popularity. If a website does not get many visitors, it will be tougher to find because not many will have it stored on their device, and eventually it might disappear altogether. On the other hand, even files that are available, can be a real pain to get if they’re large enough. You can be 80% through a download, when you find that nobody has the remaining 20%, so you can either keep it and hope that someone shows up in a week or two, or give up and look for other inserts. This is another aspect of dark nets that can be a turn-off for some – they’re very slow. My typical download speed is around 10 MB/s. While using the Tor network, it hardly ever goes above 1 MB/s, and when connecting to a hidden service it’ll rarely exceed 100 KB/s. On the Freenet, the more popular inserts feel like they hang around 50 KB/s, while the obscure files barely scratch 5 KB/s. This doesn’t feel as bad as it sounds, because the websites are often designed to be as small as possible, but regardless, do not expect high resolution audio and video, and get ready to wait a long time if you want to download anything.

There are many more networks, of course. I2P, Lokinet and the previously mentioned Gemini spring to mind. However, I cannot say anything meaningful about them because I’ve not used them. If you’d like to know more, a good place to start is Let's Decentralize which has a short summary and list of links for most of the popular networks.

Now, why would someone want to use these networks? As my explanations of Tor and Freenet have suggested, many of these networks take security very seriously. This makes them crucial for political dissidents, for example, as having their real identity exposed could put their freedom or even life at risk. It also allows the bypassing of regional censorship structures, like the Chinese Great Firewall, letting residents see content that’s been deemed wrong by their government. Most people who use these services, however, are in neither of these groups, and use them just because they value privacy and anonymity (or because the technology is cool to them). I use the Tor Browser as my primary internet browser, simply because the extra relays and standardized software makes me much less susceptible to browser fingerprinting and other methods of tracking. I don’t have anything to hide, I just don’t like Google knowing everything about me because I visited YouTube for 2 minutes. If you visit this site using that browser, a button will also appear, letting you switch to the dark net version of it – no information I post here is dangerous or illegal, I just have it set up because it’s fun.

Then, there’s the social aspect. This might sound weird, given how everyone’s supposedly trying to stay hidden and anonymous, but in fact the dark web requires much more social interaction than the web does. That’s because there are very few centralised sources of information like search engines (they do exist, but they’re quite bad) to connect users to servers. Instead, most effective ways of finding servers is by finding forums and asking people. Many of these sites are also hosted by individuals. In this respect, the dark web is quite similar to the internet from the late 90s and early 00s, colloquially known as Web 1.0, where the large, global network is separated into many smaller networks like webrings, with people exchanging links and files between one another, without central servers the likes of Reddit or Twitter. Even for the socially isolated, though, there’s still a place. Modern internet is a tool; you get on, you know what you want to do. where you’re going and how to get there. Dark web, on the other hand, encourages exploration and adventure. When you get on, you might waste an hour doing nothing productive, or you might go down a fascinating rabbit hole with websites you’ve never seen before and may never see again.

Lastly, something worth discussing is the taboo content, the illegal stuff. It’s the first thing that comes into many people’s minds when they think of the dark web. It exists, for sure, and there’s a lot of it, but not quite as much as some people may think. Rather, I suspect that people are just more bold and open about it, and that more illegal activity actually happens on the deep web. I mean, just think about how many people use drugs – do you think they all use the dark web? Most of them probably never even heard of Tor. Or how about paedophiles and child groomers? They’re more likely to hang around in places where they can actually find kids – chatrooms, private messages, Discord, Twitter and Instagram. Hell, there are many stories of people streaming torture sessions and suicides on Facebook. The and deep web that we are used to contains a lot of malicious content and people, probably much more than the dark web. The main difference comes in the approach and openness. While most people won’t show their interest in cocaine online unless they know the other person won’t rat them out, on the dark web the extra anonymity encourages them to be open, to openly discuss their experiences, recommend sellers, and so on. Different dark networks also attract different kinds of people. Tor hidden services are absolutely infested with drug stores, services offering money laundering, and so on. On Freenet, on the other hand, I’ve only seen one site where someone shared a log of their cannabis growing operation, and a handful of PDFs of books about making your own drugs, nothing on the scale of Tor. There is, however, a ridiculous amount of child pornography available on the Freenet – thankfully, the titles give you enough of an idea that you can avoid them, and besides, the more popular link indexes, like Enzo’s Index, are curated to remove these sites, and on both networks there are active communities denouncing and opposing paedophilic websites.

I hope that this text helped dispel some myths and misconceptions you might’ve believed about the hidden side of the internet. I would also be pleased if, after reading this, you’ve decided to hop onto some of these networks and check them out for yourself. If so, you can find a bunch of links for Tor and Freenet, to get started, on this website under the Other Stuff section. Happy net surfing!

Back to the top