It usually goes something like this:
''My network (or system, or application) is very secure. Periodic vulnerability scans are conducted, security patches are installed as identified, and virus detectors are implemented. Additionally, there are DMZs, firewalls, and Intrusion Detection Systems (IDS), as well as Intrusion Prevention Systems (IPS). Yep, we are totally secured. All that other policy stuff does not matter.''
Then there's this statement: ''We have all of the policies in place. We have interpreted the national and agency policies into our language and have the documents posted to our Website. Additionally, annual security training is provided to the users and the system administrators. With all these procedures and processes, how can we not be secure?''
And last, but not least, is the argument that if there is no support from senior management on information assurance, and the associated enforcement of the security policies, then the security program is doomed to failure.
Personally, I always have believed all three need to be in place to have a successful information assurance program. However, I hardly have ever seen all three areas truly working together in the same environment.
What I have noticed is the three areas are at odds with one another.
After reading a couple of recent Government Accountability Office (GAO) reports on the security posture of our government agencies, I have found I am not the only one who has noticed a lack of interoperability between the three security areas. For those of you who are not familiar with the GAO, it's an agency that works for Congress and the American people. Congress will ask the GAO to study federal programs and expenditures in an independent and nonpartisan manner. After an investigation, a report is written that identifies areas of weaknesses and provides recommendations to fix them.
One report in particular noted that the agency they were investigating was particularly weak in the implementation of the managerial and operational controls. Although the agency was diligent in periodic scanning, patch management and the conduct of technical security feature testing, the report noted that they did not track and plan for correction of their non-technical deficiencies.
The GAO report identified several specific areas...
One of the areas was risk management. According to the GAO report, the agency did not annually re-evaluate its network, system, and applications to determine residual risks. This included activities such as the security test and evaluation of the managerial and operational controls, as well as the technical controls, and the tracking and corrections of the non-technical findings.
This was particularly disturbing to me, as the determination of risk, and then its minimization, is what information assurance is all about.
The other major trouble spot was a lack of security training and awareness.
It's not that a security training and awareness program did not exist, because it does at this particular agency. The issue was that the security training program was not maintained and updated to keep pace with emerging security trends. There also was no good mechanism in place to keep track of personnel who were trained.
There were other findings as well, but these were the major points.
I have to admit that none of these findings were a huge surprise, and could probably be identified in any government agency at anytime. What did surprise me was the agency's response to the report.
Let me provide you a condensed version of the response: They periodically execute vulnerability scans, respond to the identified vulnerabilities and have other such technical controls in place. No mention of the operational or managerial controls.
I have to admit I am amazed that with all the progress we have made in the information assurance field, there could be such a lack of understanding of how the three areas must work together. However, I am hopeful that with the technical security features that are currently available for our systems; the national attention that is being given to security via policies (e.g., FISMA); and the enforcement of these security policies that is slowly but steadily being executed, we may yet get all three areas to interact in a positive manner.
Only then will our networks, systems and applications be truly secured.