Defending Critical Infrastructure Without Air Gaps And Stopgap Security

There has recently been a great amount of discussion regarding critical infrastructure and its inherent security vulnerabilities. Critical infrastructure primarily comprises aging supervisory control and data acquisition (SCADA) and industrial control systems (ICS), which are far more pervasive than most people realize: The Department of Homeland Security has defined 16 separate critical infrastructure sectors, many of which include outdated cybersecurity protections.

Security Through Obscurity No Longer Works

The vast majority of critical infrastructure consists of aging industrial control systems that were designed to operate on isolated, “air-gapped” networks. If considered at all during protocol development and network design, security took a back seat to more pressing considerations such as low latency and uptime. Multisite connectivity typically occurs via secure WAN links on private telecom networks, and operators tend to emphasize physical security over cybersecurity. Today, however, the lack of attention given to network security during early development is becoming problematic as critical infrastructure is increasingly being connected in some fashion to the Internet, giving hackers a potential access point.

Many of these SCADA and ICS systems run proprietary code on legacy operating systems that have been refined over the decades. In fact, most programmable logic controllers, protocol converters, and data-acquisition servers within these systems lack even basic authentication, making them highly vulnerable to hacking. Today, many operators believe the legacy nature of their systems confers protection, which simply isn’t true. If an asset has potential value, there are cybercriminals and nation states with the means and motives to target it.

New Thinking For The Next Generation Of Critical Infrastructure

Complicating matters further, the administrators and operations personnel tasked with supporting critical infrastructure frequently have different priorities. Operational technology (OT) teams that maintain SCADA networks focus primarily on high resiliency and availability to keep production online at any cost, while information technology (IT) teams that manage corporate networks are more concerned with connectivity, security, and compliance. However, both teams understand today’s security imperative, and within most organizations these teams are actively planning the next generation of security architectures.

As the threat landscape shifts over time, both IT and OT security infrastructure must be able to adapt to new security needs, policies, and threat-detection methods. Single-function security devices will soon be a thing of the past, as security architecture becomes increasingly versatile. Firewalls, intrusion prevention systems (IPS), VPN gateways, and routers all perform vital roles. To achieve the infrequent scheduled downtime requirements of OT environments, these software-based devices must be updatable on the fly while performing the security or networking tasks at hand. And to minimize unscheduled downtime, they must be highly reliable or support active-active clustering with transparent failover options.

In addition to support for OT protocols, it’s clear that traditional IT security solutions will need some modifications to successfully defend critical infrastructure on tomorrow’s cyber battlefields. Here’s a list of some potential features and requirements to get started:                                 

Is our industry currently providing the security technologies, flexibility, and agility to empower critical infrastructure? In many cases I believe the answer is yes, which is good news, given that many of these solutions are also required to secure the Internet of Things and the future of IT overall.