What is the deep web, and is it the same as the dark web?

More context

The internet that ordinary people picture — and that search engines map — is only a small fraction of what is actually online. The phrase deep web describes everything that exists on the regular internet but is not reachable by typing a query into Google: pages behind a login, pages behind a paywall, internal tools, private databases, machine-to-machine APIs, dynamically generated content that Google’s crawler cannot or chooses not to traverse.

The three layers, kept clear:

• Surface web. Public, indexed by Google / Bing / DuckDuckGo. Wikipedia, news sites, blog posts, marketing pages, public profiles.
• Deep web. Reachable through normal browsers, normal addresses, but not indexed because search engines cannot get past the login or paywall. Your inbox, your bank, your work tools, your hospital portal, paywalled journalism, internal company Notion, every Google Doc you have ever opened.
• Dark web. A separate, much smaller network — built on Tor, I2P, or similar privacy systems — where sites have .onion (or equivalent) addresses and only respond to those privacy networks. Accessed through Tor Browser, not your regular browser. See dark web for what actually lives there.

Why the confusion matters: media coverage routinely uses “deep web” and “dark web” interchangeably, and the word “deep” sounds mysterious. The actual deep web is the least mysterious part of the internet — it is where every account-protected, paywalled, work-internal page already lives. Your salary slip on your HR portal is on the deep web. Your child’s school grade dashboard is on the deep web. There is nothing more concerning about that than the existence of a locked office door.

The reason search engines do not index the deep web is technical and intentional. To read your inbox, a search engine would need your password — which it does not have. To read a paywalled article, the publisher would need to grant it access — which they explicitly do not. To read your internal company wiki, the company would need to publish it — which would defeat its purpose. The “non-indexed” property is a feature of every account-based service on earth.

A small overlap is worth noting: some content that was once on the surface web ends up on the deep web after deletion or password-protection, and leaked data sometimes appears both on dark-web marketplaces (where buyers find it) and on the deep web (when journalists or researchers re-publish in protected archives). That overlap is real but minor. For everyday purposes, the deep web is where almost everyone’s online life already lives, the dark web is a different network, and conflating them produces fear of the wrong thing.