Teaching network security for the SANS Institute, I am frequently asked, "Will our networks ever be secure and safe for all to use?" My simple answer is: No. We will never be able to achieve total network security. I tend to use real "doors" as an analogy to make my point. Doors have been built and refined for thousands of years. Nevertheless, burglaries occur daily, even though we know how to build safe, burglar-proof doors. With the growth of the Internet, and its increasing importance for commerce, we have to keep in mind what is a reasonable means to achieve sufficient security. Absolute security should not be the goal.
A big focus on security solutions has been patching, firewalls, virus filters, and other forms of technical security. This has been the low hanging fruit for many years, but it’s easy to exploit their vulnerabilities. Today, however, some of the most damaging exploits use "social engineering" techniques, or what is better described as "cognitive hacking," where trust becomes a casualty in an online business transaction. You don’t know that a business partnership wasn’t as trustworthy as it seemed until it is too late.
Although enhanced technology is helping to identify valid business partners in online transactions, the fundamental problem of trust in that partner still remains. The abstraction of these transactions makes it even harder to understand and trust the process. Eventually, this problem may be solved. Over time, users will better understand how online transactions work as they become more familiar with the process. Visualization and user interface technologies may make the process more transparent.
If network security problems are here to stay, then the real question is: Will these problems impede progress or even lead to abandoning network based commerce? Some scenarios predict a cataclysmic network security event leading to irreparable damage to the infrastructure. These events are sometimes described as a "super worm," assuming an automated attack that will infect and disable a large percentage of the current network infrastructure in very short time. While such an event cannot be excluded, it may be neither the most damaging nor the most likely scenario.
More damaging than a single event like a "super worm" may be the constant erosion of trust by smaller, individual events. For example, the broader scope of network security may be eroded if more people fall for simple exploits like phishing; more consumers become disappointed by ineffective and expensive anti-malware products; or a significant growth in bidders who are scammed in online auctions. For any number of reasons, current e-commerce enthusiasts may abandon the technology one at a time, adding up to a slow but fatal decline in e-commerce.
At this point, the latter scenario is not yet a reality, but it’s a possibility. Efforts to push more responsibilities to the consumer and continued use of inadequate authentication techniques will lead to more disillusionment of online security. When you enter a store, you expect sincerity, reliability, and competence from the sales clerk. These principles must be applied to online commerce in order to build consumer trust. However, with the hasty implementation of online commerce, competence is missing in many cases. To achieve a level of sufficient network security, lessons learned from the real world cannot be forgotten.
— Johannes Ullrich, PhD, Chief Research Officer, SANS Internet Storm Center