Computing at the edge – cybersecurity overstretched?
Article information and share options
Edge computing can supercharge data exchange, but also lead to more cybersecurity breaches.
Fast broadband connections and central cloud servers enable the rapid transfer and process of massive amounts of information on the data highway. However, with Internet of Things (IoT), data needs to be transferred and processed not just more quickly, but instantaneously. Think of autonomous vehicles: any time-lag in signal transmission and processing can prove fatal.
This is where edge computing comes in. To minimise latency in data transactions, computing power is added close to the connected end-devices themselves. In other words, at the periphery or edge of a network. Edge computing does not replace cloud services. It complements them by transferring the processing power from cloud platforms to where the data is created and consumed. Edge computing is playing a pivotal role in innovating and maintaining digital ecosystems across manufacturing, utilities, robotics and all other spheres demanding low-latency, a development in which 5G is likely to play an ever-increasing role.
The addition of more and ever-evolving interconnected devices to a network also, however, increases potential attack surfaces.
Poor implementation of edge computing can expose system vulnerabilities, which hackers could target using latest innovations in artificial intelligence to search through codes for entry points and deploy intelligent malware. Decentralised by nature, edge computing is less likely to benefit from a strong security monitoring. Data collected, consumed and intermediately stored at the edge can also be vulnerable to environmental conditions, and these can be harsh, such as in, for instance, the context of a wind turbine or in agriculture. This adds physical challenges to the maintenance of devices as well as data retention and security.1 By moving security concerns to the periphery, edge computing heightens cyber risk potential from under-service, negligence and blind spots. Any device that remains connected to a network beyond its projected life span and is no longer updated with adequate security patches invites attackers. A scenario of “internet of forgotten things” that are more vulnerable to cybercrime is not a long stretch of the imagination.2
- Edge computing leads to higher cyber exposures in industry and other sectors. There will also be exposures in professional and consumer solutions involving mobile devices.
- The increased complexity of systems/ networks with the addition of edge computing could increase potential cyber risk exposures.
- Decentralisation may also favour lower severity but higher frequency risk.
- Edge computing often takes place at remote environments and under conditions of limited physical security. Cyber incidents can cause machine failure or malfunction, and even business interruption. In connection with autonomous vehicles or health-critical devices, this can lead to injuries or fatalities.
- Liability may be more difficult to assign in case of failure in the edgecomputing world.
1 iSF 30 Information Security Forum, Threat Horizon 2022: Digital and physical worlds collide, iSF 30 Information Security Forum, Jan. 2020.
2 Ibid., p 25ff.