The Internet is a global system of interconnected computer networks that communicate using standardised protocols, principally the Internet Protocol suite. It is not a single network owned or operated by one organisation, but an interconnection of independently administered networks known as autonomous systems. These networks include Internet service providers, telecommunications operators, universities, enterprises, governments, content providers, cloud platforms and other organisations. The Internet enables hosts and networks using different physical technologies to exchange packets through common protocols. Its decentralised structure is one of its defining characteristics: there is no single central machine that constitutes the Internet and no single organisation that controls all Internet traffic.
The Internet developed from earlier packet-switching research and networking projects, particularly the ARPANET, which began operational development in the late 1960s. The transition to TCP/IP was a major milestone because it allowed independently operated networks to interconnect using a common internetworking architecture. On 1 January 1983, ARPANET formally transitioned to TCP/IP, a date often treated as a major landmark in Internet development.
The Domain Name System was introduced later to provide a distributed mechanism for mapping human-readable domain names to network addressing information. Commercial Internet access expanded significantly during the 1990s as Internet service providers, businesses and consumers adopted the technology. The Internet subsequently became a fundamental communications infrastructure for modern computing.
The Internet's basic operation depends on layered protocols. At the network layer, IP provides addressing and packet forwarding between networks. TCP provides reliable, ordered, congestion-controlled byte-stream transport, while UDP provides a simpler datagram transport without TCP's reliability mechanisms. Application protocols such as HTTP, SMTP, DNS, SSH and others operate above these transport mechanisms. Ethernet, Wi-Fi, fibre-optic links, cellular radio and satellite links can all carry Internet traffic because the higher-layer protocols abstract away many details of the underlying physical medium. This separation is a central reason why an application can communicate across heterogeneous infrastructure without needing to understand the exact physical path taken by every packet.
Internet routing is primarily based on the Border Gateway Protocol, or BGP, which exchanges reachability information between autonomous systems. Within an organisation, routing protocols such as OSPF or IS-IS may calculate paths through the internal network, while BGP handles inter-domain routing between separately administered networks. Routers examine destination IP addresses and use routing tables to determine where packets should be forwarded. The path chosen is not necessarily geographically shortest because routing decisions incorporate administrative policies, commercial relationships, topology and routing attributes. Consequently, two packets between the same endpoints can potentially follow different paths, and Internet paths can change when links fail, congestion occurs or routing policies change.
The Internet supports a vast range of services, including the World Wide Web, electronic mail, voice and video communications, cloud computing, software distribution, online gaming, financial transactions, remote administration and machine-to-machine communication. These services are applications that use Internet connectivity rather than being synonymous with the Internet itself. This distinction is particularly important because the Web is only one service operating over the Internet. Email, for example, commonly uses SMTP for message transfer and other protocols for retrieval, while the Web uses HTTP or HTTPS. A mobile application may communicate with an Internet server without using a web browser or the World Wide Web at all.
Security on the Internet is implemented through multiple layers rather than by IP itself providing complete protection. TLS can provide encryption and authentication for application connections such as HTTPS. Firewalls can filter traffic according to addresses, ports, protocols and other attributes. Virtual private networks can create encrypted tunnels across untrusted networks. DNS security mechanisms, authentication systems, endpoint protection, intrusion detection and secure application design address additional threats. Internet-connected systems face risks including malware, phishing, denial-of-service attacks, credential theft, routing attacks, software vulnerabilities and data interception. The openness and global reach of the Internet provide substantial benefits while also exposing connected systems to traffic from potentially untrusted networks.
Internet governance is distributed among technical organisations, standards bodies, network operators, governments, companies and other stakeholders. Organisations such as the Internet Engineering Task Force develop and document many Internet protocols through the Request for Comments process. ICANN coordinates important global Internet identifiers, including domain-name systems and certain aspects of IP address allocation through regional Internet registries. Governments can regulate Internet services within their jurisdictions, while network operators establish their own technical and commercial policies. This produces a governance model that is neither entirely centralised nor entirely unregulated. Standards interoperability remains essential because independently operated networks must agree on common protocols for global communication to work.
A common misconception is that the Internet and the World Wide Web are interchangeable names. The Internet is the underlying global internetwork and associated protocol ecosystem, whereas the Web is an application platform built on top of Internet connectivity. Another misconception is that packets always travel through one fixed route between two devices; Internet routing is dynamic and policy-driven.
The Internet also does not guarantee that every packet will arrive, that every route will be optimal or that every application will be secure. IP provides a mechanism for internetworking, while reliability, confidentiality, authentication and application-level functionality are supplied by other protocols and systems. Understanding these distinctions is essential when analysing Internet architecture and diagnosing network behaviour.
