The Washington D.C.-based think tank has installed CounterStorm-1, a new network security appliance designed to detect and stop zero-day attacks. CounterStorm, Inc., a New York City-based security company, just announced the product release, though it has been deployed to 10 different corporate and government environments over the last month.
''Zero-day attacks haven't been a problem for us yet, but this alerts us if something different is happening on the network,'' says Jane Fishkin, CIO and vice president of technology at The Brookings Institution.
Zero-day attacks are defined as malicious exploitation of previously unknown vulnerabilities before patches and signatures are available. While not a prevalent problem at this point, most security advisors say it's only a matter of time before they become a serious issue.
''It's an insurance policy that has paid off a number of times,'' says Jake Marshak, director of network and system infrastructure at the institution, which runs 600 desktops and laptops, along with 50 servers. ''It's not the end-all and be-all of our security, but it's a piece of that. It's looking at our traffic patterns in a unique way and looking for anything anomalous, and then it responds. It's not just an alerting package but a responding package.''
And the fact that the product detects and responds to threats is a big deal for Marshak.
''That's huge for us,'' he says. ''We're always busy doing other things. Our staff isn't in a position to always be watching a monitoring console. To have it take the first steps of response is a requirement. It wouldn't be very helpful without that.''
Most anti-virus companies issue signatures or virus updates to protect networks against incoming threats. When a software vendor learns of a vulnerability in its software, its developers and analysts create a patch to fix it.
The danger with zero-day attacks is that the attack comes before a patch can be released and before anti-virus signatures can be sent out. The attack might even come before the IT and security communities even know the vulnerability exists.
So how can an IT or security administrator fend off that kind of threat?
Matt Miller, vice president of engineering at CounterStorm, says the key lies in not relying solely on signatures and patches.
He explains that CounterStorm-1 uses a combination of behavioral attack recognition, anomaly detection and a dynamic honeypot. Information from each of these three engines is correlated in real-time to enable containment of malicious activity.
''The signatures and patches are still not coming very quickly,'' says Miller. ''A really good vendor will take five to eight hours to issue a signature or patch.'' Miller adds that one of their customers was hit by a zero-day attack this past summer. An e-commerce server was penetrated and information was being culled and sent to China. The attack was made taking advantage of a vulnerability that wouldn't be announced for several more days.