A private network is one that is logically (and sometimes physically) separate from the public internet. Traffic on a private network stays within a defined boundary, typically reaching only systems that have been explicitly authorised to participate.

In short: For cellular IoT, a private network usually means a combination of private APN (the SIM connects to a dedicated access point within the mobile operator's network rather than the standard public internet APN), private IP addressing (devices are assigned addresses from RFC 1918 ranges that are not routable from the internet), and a defined breakout point (the operator's private network connects to the customer's network via a dedicated link, VPN tunnel, or peering arrangement rather than via the public internet).

Private networks matter for IoT for both security and regulatory reasons. The Network and Information Systems Regulations 2018 (NIS), the incoming UK Cyber Security and Resilience Bill, PCI DSS for payment connectivity, and operator-specific frameworks for critical national infrastructure all push toward private networking for connected devices that handle sensitive data or support essential services.

The trade-off with private networks is cost and complexity. A standard public-internet IoT SIM is cheaper than a private APN SIM, and easier to set up. For deployments with low risk profiles and modest data volumes (a temperature sensor on a non-critical asset), public-internet routing with good device-level security is often appropriate. For critical infrastructure, payment connectivity, and sensitive operational technology, the cost of a private network is usually justified by the reduced risk and easier regulatory compliance.