Georgetown Conference Highlights Cybersecurity Challenges
"Spear-phishing" and BYOD were among the biggest concerns, as enterprises try to re-imagine their models for keeping data and communications secure.
Most continuing legal education seminars are dull affairs involving lawyers and "talking head" lectures. But last week's Cybersecurity Law Institute at the Georgetown University Law Center was no ordinary event; not only did it get the Wall Street Journal's attention, it attracted an unusually diverse group of attendees, including CIOs, CISOs, and government officials. The cornerstone of the two-day seminar was a lively simulation of the C-suite struggling to cope with a recently-discovered data breach that graphically highlighted the challenges of managing such a breach.
Here are my impressions from the two-day conference:
The Situation is not Hopeless, and Doing Something is Better than Doing Nothing: While much of the seminar focused on what can go wrong, Tony Sager, formerly of the NSA and now Director of the SANS Institute, observed that most security problems involve known problems that have known solutions. But not everybody who needs help knows what those solutions are. Moreover, many enterprises are unsure of what concrete steps they should take to increase cybersecurity.
A good starting point for enterprises is to begin implementing the 20 Critical Security Controls (the "CSCs"), a step that will help to manage the 20% of the issues that cause 80% of the damage (Sager noted that Verizon's annual Data Breach Investigations Report now maps to the CSC in an effort to help enterprises understand the utility of the individual CSCs).
Planning and Practice: Another theme for enterprises was how cybersecurity planning could benefit from tracking key aspects of business continuity and disaster recovery (BCDR) plans. As with BCDR, enterprises shouldn't wait until the crisis hits before developing cybersecurity plans. And those plans need to be tested and refined periodically. One panelist summarized his key takeaway from the conference as follows: "I'm going back and do tabletop exercises with my cybersecurity plans every chance I can."
Fear the Spear: Security experts, CIOS and CISOs agreed that spear phishing is an increasingly grave threat. Former FBI agent Shawn Henry, now of Crowdstrike Services, observed that the sophistication of spear phishing attacks make them almost impossible to prevent. As a result, training and maintaining employee awareness of the dangers becomes critical. Northrop Grumman's Michael Papay spear phishes his own employees to keep them alert. For example, on April 12 he sent a bogus "IRS email" to 68,000 of the company's employees regarding their tax returns that succeeded in fooling several people.
While training employees and building awareness is important, some spear phishing attacks will succeed. As a result, enterprises must develop the capacity to limit an attacker's ability to exploit a breach--by, for example, blocking outbound traffic to new and uncategorized websites.
Can BYOD and Security Co-Exist? While wireless security was not a focus of the conference, it was clear that some security-minded enterprises are wary of the risks of a BYOD model. For example, neither Lockheed Martin nor Northrop Grumman allows personal wireless devices on their network. Interestingly, one Northrop Grumman attorney told me that while she could access email on her personal device using a containerized solution, her strong preference was to keep personal and business devices separate.
This anecdote suggests that security may stymie BYOD in industries where security is paramount (e.g. financial services, health care and defense). It might also cause enterprises in other industries to revisit their mobile strategies given the considerable advantages of corporate-liable wireless programs.
Enterprises Need Eyes and Ears to Improve Security: Mandiant's Shane McGee noted that the victim discovers data breaches only 6% of the time; the other 94% of the time they find out from a third party such as the FBI or a service provider. Worse, it takes an average of 416 days for a corporation to discover that security has been compromised.
This puts a premium on developing the capability to detect and respond to security threats. While some carriers used to provide fraud warnings when they detected unusual traffic patterns (as in the case of PBX hacking), that practice has waned, and it's likely that any notice that does come from a sympathetic law enforcement agency or an ISP will be frustratingly vague.
As a result, customers who want to enlist the expertise of their carrier or ISP in detecting threats have to purchase separate security services. U.S. Magistrate Judge John Facciola sounded a cautionary note about enterprise contracting practices, saying that he was "astonished" at how many sophisticated parties fail to read and understand the terms and conditions of their cloud and other agreements that have a direct bearing on cybersecurity.
Communication is Key: The importance of communication came up repeatedly during the conference. Participants in the data breach simulation emphasized the need for smoothly-functioning communication channels both within the enterprise and between cybercrime victims and law enforcement. Meanwhile, Tony Sager of SANS Institute characterized cybersecurity as an information management problem that focuses on getting the right information to people who need it.
The Importance of Mental Models: A lingering question from the conference concerns our mental models for cybersecurity. The reigning model is the "citadel paradigm," in which the enterprise is a fortress protected by "electronic walls and a moat" of firewalls, antivirus software and other security products. Indeed, the data breach simulation at the conference was firmly rooted in a variant of the citadel model: the corporate ship had struck an iceberg, and the participants were trying to plug the leak to avoid catastrophe. But what (to continue the nautical metaphor), if the main challenge for the enterprise/ship isn't avoiding stray icebergs, but coping with dozens of persistent, small leaks that could overwhelm the pumps?
The demise of the "citadel paradigm" was chronicled in a recent New Yorker article by John Seabrook entitled "Network Insecurity." Seabrook quoted one security expert who said that, up until four years ago the citadel paradigm remained viable, and "encryption and firewalls were doing a pretty good job. But the latest attack kits are bypassing those perimeter defenses, which is why this paradigm has to shift."
There are several alterative paradigms that might be more useful. One panelist at the conference told me during a break that he favored analogizing cybersecurity to policing a large city: preventing all crime is unrealistic, and different "neighborhoods" will face different threats. With any luck, next year's Georgetown cybersecurity conference will work to develop a useful paradigm to help guide future enterprise cybsersecurity efforts.