Vulnerabilities / Threats

10:30 AM
Rick Howard
Rick Howard
Connect Directly
E-Mail vvv

The End Of Whac-A-Mole: From Incident Response To Strategic Intelligence

In the face of mounting cybercrime, hacktivism, and espionage, network defenders need to transform their tactical IR groups into full-scale cyber intelligence teams.

I am an old U.S. Army guy. In the early 2000s, I ran the Army’s Computer Emergency Response Team, the ACERT, and my job was to coordinate offensive and defensive operations across the Army. One of my main tasks back then was to respond to computer incidents.

We had teams of incident responders that would wait for an incident to happen, spend weeks trying to determine the exact location of the infected endpoint somewhere in the Army network, travel to said endpoint once the location was known, take the machine off the network, perform some rudimentary forensics on it (because that was all we knew back then), re-image the endpoint, put it back on the network, and wait for the next incident to happen. Even back then, I affectionately referred to that process as playing Whac-a-Mole with the adversary. But back then, Whac-a-Mole is what most network defenders were doing.

Fifteen years later, advanced network defenders have realized that playing Whac-a-Mole with the adversary is not sufficient. It is lacking in several key areas. Whac-a-Mole gives the network defender no intelligence about what these adversary groups were trying to accomplish, why they were trying to accomplish it, where they went once they compromised the endpoint, and whether or not they had the capacity to materially impact their business or organization. Playing Whac-a-Mole with the adversary back in the early 2000s was fine because we did not know any better. Today, the process seems quaint and is very tactical.

Since then, advanced network defender organizations have realized that they need to up their game. We all need to transform our tactical incident response teams into strategic intelligence teams. Here is what I mean.

Adversary groups are made up of people. They have motivations like crime, hacktivism, espionage, and warfare. Successful adversary groups design and execute campaign plans against a target, as opposed to making it up on the fly. In a way, they are similar to the network defenders in the world. Just like us, they do not have unlimited resources and, just like us, they are not superheroes. They put their pants on the same way we do.

Wanted: full-scale intelligence teams
The successful adversary group develops repeatable processes so that it does not have to re-invent the wheel every time it goes after a new victim. Those repeatable processes leave traces in the network. We call them “indicators of compromise” and we can track them. In my day, we assigned this work to the tactical incident responders, but it became clear early on that those groups did not have the skillset needed to track adversary groups with this kind of granularity. We still needed the incident responders to do what they do, but it was not sufficient to gain an understanding of what the adversary group was about in the aggregate. We needed full-scale intelligence teams.

In contrast to an incident response group, an intelligence team collects raw information about adversary groups, transforms that raw information into intelligence through analysis – something a leader can use to make a decision – and then disseminates that intelligence to the people in the organization who can use it. This is called the intelligence life cycle. The U.S. military started to build these new cyber intelligence teams in the early 2000s to track Chinese government-sponsored cyber espionage under a code name called Titan Rain.

The commercial world started to embrace the idea of intelligence teams when the Titan Rain adversary group started hitting the Defense Industrial Base (commercial companies that sell goods and services primarily to the U.S. government, like Raytheon, Lockheed Martin, Boeing and Northrop Grumman) in 2003. But the traditional commercial world –  companies that are not part of the Defense Industrial base (DIB) – really did not embrace the idea until 2010, when Google got hit by an advanced cyber espionage adversary group and publically called the government of China out as being the perpetrator of the attack.

Since then, advanced organizations in the non-DIB space have been slowly warming to the idea that they need strategic cyber intelligence teams in addition to their tactical incident response teams. Some groups have even combined the missions. I believe that if you don’t have an intelligence team in your organization – even if it is just a team of one person – you cannot hope to keep up with the advanced adversary.

The evolution has begun. Advanced organizations have already transformed their tactical incident response teams into strategic intelligence teams. Palo Alto Networks created their team, Unit 42, last year. You are going to have to do this eventually. I recommend that you start now.

Rick Howard is Chief Security Officer for Palo Alto Networks, where he is responsible for internal security of the company as well as developing the Threat Intelligence Team to support the next-generation security platform. He previously served as Chief Information Security ... View Full Bio
Comment  | 
Print  | 
More Insights
Newest First  |  Oldest First  |  Threaded View
User Rank: Moderator
7/26/2015 | 8:57:20 PM
No staff for this...
I agree in principle with your idea.  However, over the past 14 years I have managed SecOps teams in Defense, Energy, and Finance companies and not once did we have the minimum number of staff resources necessary to do our basic jobs.  Based on that real life experience I see no way typical organizations will have the necessary staff to pull this off.  And Please Please Please do not say theres a 3rd party company that can help with that.  In the famous words of Merlin "Never again!!!"

Anyway, this lack of staffing contiunues to push more and more organizations in finding and implementing auto-blocking technologies and real time threat intelligence & monitoring systems like Splunk ES to compensate for the staff shortages.  So no, what you propose is a great idea but that's it.  Until organizations properly invest in the human capital needed in cybersecurity we will have no real choice but to continue to play whack-a-mole.
White House Cybersecurity Strategy at a Crossroads
Kelly Jackson Higgins, Executive Editor at Dark Reading,  7/17/2018
Mueller Probe Yields Hacking Indictments for 12 Russian Military Officers
Kelly Jackson Higgins, Executive Editor at Dark Reading,  7/13/2018
10 Ways to Protect Protocols That Aren't DNS
Curtis Franklin Jr., Senior Editor at Dark Reading,  7/16/2018
Register for Dark Reading Newsletters
White Papers
Cartoon Contest
Write a Caption, Win a Starbucks Card! Click Here
Latest Comment: This comment is waiting for review by our moderators.
Current Issue
Flash Poll
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
PUBLISHED: 2018-07-19
redhat-certification does not properly restrict files that can be download through the /download page. A remote attacker may download any file accessible by the user running httpd.
PUBLISHED: 2018-07-19
redhat-certification does not properly sanitize paths in A remote attacker could use this flaw to overwrite any file, potentially gaining remote code execution.
PUBLISHED: 2018-07-19
The approveAndCall function of a smart contract implementation for Aditus (ADI), an Ethereum ERC20 token, allows attackers to steal assets (e.g., transfer all contract balances into their account).
PUBLISHED: 2018-07-19
TP-Link WR840N devices allow remote attackers to cause a denial of service (connectivity loss) via a series of packets with random MAC addresses.
PUBLISHED: 2018-07-19
AVEVA InduSoft Web Studio v8.1 and v8.1SP1, and InTouch Machine Edition v2017 8.1 and v2017 8.1 SP1 a remote user could send a carefully crafted packet to exploit a stack-based buffer overflow vulnerability during tag, alarm, or event related actions such as read and write, with potential for code t...