Does the constant connectivity of IoT devices open up more doors for cyber attacks?

Connection design, permissions and maintenance determine exposure
Hello, friends. Yes, continuous connectivity can create more opportunities for cyberattacks, but being online does not automatically make a device accessible to everyone. In Internet of Things systems, assess exposed interfaces, command permissions, software condition and the effect on connected systems together. Exposing the same device directly to the internet differs from allowing it to reach specific services through a restricted network.
1. Why does continuous connectivity matter?
While a vulnerable, reachable service remains online, opportunities for automated scanning or unauthorised access attempts can persist. The attack surface includes network services, management interfaces, apps, cloud accounts and connected services. However, there is no universal calculation that a device online for 24 hours carries twice the risk of one online for 12. Access controls and the nature of vulnerabilities change the outcome.
Assess internet connection requirements by function. A device can operate continuously on a local network without internet access, while another may contact the cloud only periodically. Local-network connectivity, outbound internet access and direct reachability from the internet are different conditions.
2. Devices may be reachable through several paths
With direct access, a management or communication service is exposed externally. With cloud-mediated access, the device connects to a provider and the app uses that service. With local access, another device or user on the same network is involved. Closing one path does not automatically close the others.
NAT on a router or the absence of port forwarding does not provide complete IoT security by itself. A device's outbound session, a compromised cloud account or a weakness on the local network may still be misused. Assess IPv4 and IPv6 access rules, cloud accounts and remote-access features together according to the deployment.
Editor's note: “No externally exposed port” does not mean “no remote command path”. If a phone can control the device through a cloud service, a remote-control path exists. That can be useful; its security depends on account protection, service authorisation and the commands accepted by the device.
3. Stealing information is not the only possible objective
Reading information, changing settings, issuing commands, disrupting service and recruiting a device for other attacks are different outcomes. Functionality can be affected without a data breach. A camera could continue displaying images while generating unwanted traffic; a sensor could appear online while returning stale readings. These are possible scenarios, not evidence that every connection problem is an attack.
4. What did the Mirai example demonstrate?
A US Department of Justice announcement dated 13 December 2017 describes Mirai compromising connected cameras, routers and recording devices and using them in DDoS attacks. A botnet is a group of compromised devices controlled together. A DDoS attack aims to disrupt a target service through traffic or requests from many sources. This historical example shows that a device may have value to an attacker beyond its owner's information.
5. Compromise does not automatically spread to everything
One compromised device does not inevitably lead to control of the entire network. Propagation depends on network reachability, shared credentials, exploitable weaknesses and permissions. Separating IoT equipment from critical computers and restricting communication between them can reduce potential consequences. A separate SSID or VLAN with unrestricted routing does not deliver the intended protection.
Wi-Fi security protects the wireless connection but does not replace cloud-account or application security. Similarly, hubs and account links added for device compatibility introduce access relationships. Define clearly which systems may only read information and which may issue commands.
6. What does restricting communication achieve?
A device with a defined purpose often has communication requirements that can be limited. Defining necessary destinations, services and directions reduces unnecessary access. The MUD approach in IETF RFC 8520 lets a manufacturer describe expected device communication. It is not a standard feature of every product; compatible devices and network infrastructure are required. It also does not replace security patches.
For example, a hypothetical environmental sensor that only reports readings may not need access to every office computer. However, indiscriminate blocking can disrupt updates, time synchronisation and alarms. Test required functions after applying rules rather than checking only whether connectivity has been blocked.
7. What do strong passwords and encryption leave unresolved?
Unique passwords, appropriate multifactor authentication and correct certificate validation reduce access risks. However, compromising an overprivileged account can allow apparently valid commands over an encrypted connection. Reusing one credential across all devices can increase the impact of a single incident. Manage credentials individually where possible and revoke unused accounts and connections.
8. Can continuous connectivity also improve security?
Secure updates, incident notifications and remote health monitoring can depend on connectivity. Cutting off maintenance and updates entirely may make known vulnerabilities harder to fix. Simply disconnecting everything is therefore not appropriate for every use. Track vendor support, the running version and update provenance. IETF RFC 8576 considers IoT security challenges throughout the device lifecycle.
Limited processing power alone does not establish insecurity; design and capabilities matter. Access restrictions may provide an interim measure for an unsupported, unpatchable device, but a replacement or safe-isolation plan is needed. Include these requirements in long-term costs.
9. How do you distinguish an attack from a connection fault?
Traffic to unexpected destinations, repeated failed logins, changed settings and unusual restarts warrant investigation but are not proof of an attack individually. Power problems, wireless coverage and service outages can produce similar symptoms. Normal behaviour, accurate timestamps and appropriate logs provide a basis for comparison.
Preserve relevant records, restrict risky access and revoke compromised credentials when investigating an incident. Restarting alone may not resolve the cause. Monitoring records also require protection of personal information. Cybersecurity monitoring does not justify unlimited surveillance of users.
10. Industrial and critical applications need additional consideration
The difference between IoT and IIoT is especially relevant here. Industrial cybersecurity must account for safe process behaviour. Collecting information through a Modbus-connected gateway should not require unnecessary write access to the control network. Design monitoring permissions and control permissions separately.
Disconnecting a suspect device's power or network without considering its function can affect alarms or control. The consequences of interruption also need to be established for healthcare applications and other critical functions. NIST SP 800-82 Rev. 3 considers OT security alongside reliability and safety. Internet outages, cloud-service outages and power failures are different conditions; test which local functions survive each one.
11. How can you verify the installation?
In a hypothetical facility, checks could confirm that a sensor reports only to an approved hub, users have the intended permissions and a stale-data warning appears when communication fails. A connected indicator does not prove that a reading is current. Authorised personnel should plan network and permission checks; avoid unplanned scanning or experimentation on production equipment.
12. Which connections should remain enabled?
Identify connections serving functions that are actually necessary, and remove unused remote access, old accounts and integrations. During transfer or electronic-waste disposal, remove cloud associations and credentials too. Effective protection combines controlled necessary connectivity, minimal permissions and a prepared response to problems.
Related guides
Fundamentals: IoT guide, How IoT works and IoT versus IIoT.
Technical protection: IoT security, Maintenance and updates, Data-breach risks and Compatibility issues.
Use and information: Privacy and personal data and Long-term costs.
Sources
NISTIR 8228: IoT cybersecurity and privacy risks
NIST SP 800-82 Rev. 3: operational technology security
IETF RFC 8520: MUD and restricting communication to intended use
IETF RFC 8576: IoT security challenges and lifecycle
ETSI EN 303 645 V3.1.3: consumer IoT security and attack surfaces
FTC: secure IoT design, access control and ongoing support
US Department of Justice: Mirai and DDoS, announcement of 13 December 2017