A dark net (or dark internet) is an overlay network that can only be accessed with specific software, configurations, or authorization, often using non-standard communications protocols and ports. Two typical dark internet types are friend-to-friend networks (usually used for file sharing with a peer-to-peer connection) and privacy networks such as Tor.
If we conceive of the Web as a data ocean, most of us are interacting with the wavy, transparent, easily navigable Surface Web. The Surface Web is the portion of the Web that has been crawled and indexed (and thus searchable) by standard search engines such as Google or Bing via a regular web browser.
In the darkness below, beneath the electronic thermocline, are the abyssal depths of the Deep Web (also referred to as the Invisible Web or Hidden Web) – the portion of the web that has not been crawled and indexed, and thus is beyond the sonar reach of standard search engines. It is technically impossible to estimate accurately the size of the Deep Web.
However, it is telling that Google – currently the largest search engine – has only indexed 4-16 percent of the Surface Web. The Deep Web is approximately 400-500 times more massive than the Surface Web. It is estimated that the data stored on just the 60 largest Deep Web sites alone are 40 times larger than the size of the entire Surface Web.
[Click on the image to zoom]
Growing rapidly within the Deep Web is the Dark internet (also referred to as the Dark Web, Dark Net, or Dark Internet). Originally, the Dark internet referred to any or all network hosts that could not be reached by the Internet.
However, once users of these network hosts started sharing files (often anonymously) over a distributed network that was not indexed by standard search engines, the Dark internet became a key part of the Deep Web. Unlike the traffic on the Surface Web or most parts of the Deep Web, most Dark net sites can only be accessed anonymously.