Network segmentation is not a new concept. It’s been applied to improve performance by reducing collision domains in shared Ethernet networks and broadcast domains in Layer 2 networks. However, the idea of network segmentation as a cybersecurity measure is still relatively young. In years past, it was sufficient to protect the perimeter of a network to keep undesirable traffic out, and if that effort was successful, the internal network was deemed to be trustworthy. The network effectively had a hard, crunchy shell, but was soft on the inside.
Fast-forward to today, and we still see the perimeter-centric security model in practice. However, malicious actors have become experts at circumventing such basic controls. Recent breaches to organizations in the financial services industry have led to fraudulent international fund transfers, jackpotting of ATMs infected with malware delivered via the bank’s internal network, and theft of PII by or through third-party partners. In several examples, after the initial compromise, the open internal network allowed the attackers to move freely through the targeted environments in search of valuables. It’s time to improve security posture using internal network segmentation to limit exposure and contain insider threats.
Without hesitation, anyone building a brand-new enterprise network should incorporate segmentation gateways into the design. A full-blown Zero Trust network would be even better to protect critical data from unauthorized access, reduce exposure of vulnerable systems and limit lateral movement of attackers. However, few network architects and engineers have the luxury of starting from scratch. For the most part, everyone must work with a legacy network that is soft inside and uses traditional technology. So, when opportunities to redesign a portion of the network arise, seize them to introduce some degree of network segmentation. For example, projects to introduce private cloud computing or redesign the remote office network may be leveraged to compartmentalize that piece of the environment. Public or hybrid cloud computing efforts also present opportunities to include network segmentation principles in those designs.
Several of our financial sector customers have seen a greater interest in network segmentation from their regulators. In one instance, a multi-national bank segmented “high-risk” countries from the rest of its internal network at the suggestion of its regulator. The respective regulators for two other financial institutions inquired about their plans for network segmentation. In May 2017, one of the U.S. federal banking regulators, the Federal Deposit Insurance Corporation (FDIC), was cited for insufficient isolation of financial systems from other parts of its own network. It should come as no surprise if the FDIC or other banking regulators place a greater emphasis on network segmentation efforts when they next examine their supervised institutions. This trend is not isolated to the U.S. – our customers in other regions have also noted more regulatory interest in network segmentation as a cybersecurity measure.
The reality is that many financial institutions still find themselves with open internal networks. Multiple challenges, including limited knowledge of applications in their environments and the potential for business disruptions, have slowed the adoption of network segmentation for security purposes. However, obtaining visibility into the actual application traffic on the network, along with a practical and measured approach to segmenting the internal network, will minimize the likelihood of disruptions to business. In the end, adding some amount of network segmentation will create another layer of defense and reduce the potential for data exposure after an intrusion.
Learn more about network segmentation for the financial sector and the Zero Trust model: