Consensus is growing for the reform of flawed FISMA

Congress is contemplating changes to the Federal Information Security Management Act of 2002, and a panel of government and private-sector experts provided some strong arguments for doing so when they testified before a House subcommittee March 24.

Their assessment was not entirely negative. “The passage of the Federal Information Security Management Act in 2002 served as a game-changing event for the federal agency community,” said John Streufert, the State Department's chief information security officer. In addition, Gary Guissanie, acting deputy assistant secretary of Defense for cyber identity and information assurance, said the eight-year-old law has “significant strengths in improving cybersecurity.”

But the witnesses who testified before the House Oversight and Government Reform Committee's Government Management, Organization and Procurement Subcommittee agreed that the results have been disappointing.

“Despite the improvement reported by agencies, the federal government’s communications and information infrastructure is still far from secure,” said federal chief information officer Vivek Kundra. “The FISMA measures reported on annually have led agencies to focus on compliance. However, we will never get to security through compliance alone.”

What the experts said is needed is a shift away from what critics called a culture of compliance to a performance-based model that focuses on continuous monitoring and managing real risks. Kundra and subcommittee Chairwoman Rep. Diane Watson (D-Calif.) cited the State Department as an example of how that approach could improve information security.

Watson has introduced a bill that would amend FISMA to codify many of the necessary changes. The Federal Information Security Amendments Act of 2010 (H.R. 4900) would create a National Office for Cyberspace at the White House to oversee the nation's cybersecurity posture. It would guide policy and standards; require agencies to perform automated, continuous system monitoring and penetration testing; require independent auditing of the programs' effectiveness; and incorporate security requirements into acquisition policies.

Subcommittee member John Duncan Jr. (R-Tenn.) said he was skeptical that the government would ever be able to secure its information systems and was beginning to suspect that investing in security was like “throwing money down a rat hole.”

That would be a big rat hole. Last year, agencies reported spending $6.8 billion on information security out of about $75 billion spent on IT overall, according to the subcommittee.

However, the fact that such a figure is even available reflects the changes taking place in federal cybersecurity. Last year, the Office of Management and Budget, which oversees FISMA compliance, began requiring agencies to provide information on security-related costs. In an effort to improve the monitoring of cybersecurity efforts, OMB also implemented Cyberscope, an automated, interactive reporting tool that replaces manual processes.

But problems remain. Gregory Wilshusen, director of information security issues at the Government Accountability Office, said agencies "continue to tread water in securing their systems.”

“An underlying cause for information security weaknesses is that [agencies] have not yet fully or effectively implemented key elements of an agencywide information security program,” he said.

John Gilligan, who was the Air Force's CIO when FISMA was enacted and is now a consultant, said FISMA's implementation is to blame.

The emphasis on adopting an impossibly large catalog of security controls without focusing on high-priority items has led to what he called a scatter-shot approach to security. Furthermore, Congress' grading of FISMA compliance has emphasized characteristics that are easily measured but have little correlation to actual security. Consequently, staff and budget resources have been eaten up by reporting processes that have done little to improve security.

“In my view, the implementation of FISMA has been like getting on a treadmill,” he said. “A treadmill is great if all you want is exercise, but it is not the way to reach a destination. The federal government has certainly burned a lot of calories, but we are still a long way from reaching our destination of dramatically improved security.”

Alan Paller, director of research at the SANS Institute and a longtime critic of FISMA, was less generous, saying that FISMA has impeded the security process and wasted billions of dollars.

Streufert outlined steps the State Department has taken to counter increasing threats, which include 3.5 million spam e-mail messages, 4,500 viruses and more than 1 million external network probes each week.

“In fiscal 2009, the department began supplementing FISMA compliance reports and studies with a risk scoring program, scanning every computer and server connected to its network not less than every 36 hours on eight security factors and twice a month for safe configuration of software,” he said.

The program uses the Consensus Audit Guidelines to assess critical security controls and the National Vulnerability Database to assess vulnerabilities. Configuration scans are performed every 15 days, problems are assessed daily, and letter grades are issued monthly to senior managers. In the first year of scoring, the overall risk to the department's key unclassified network was reduced by nearly 90 percent.

“These methods, however limited, have allowed one critical piece of the department’s information security program to move from the snapshot in time previously available under FISMA to a program that scans for weaknesses on servers and personal computers continuously,” Streufert said.

Kundra said improving the government’s security posture would require:

  • Enhancing coordination among agencies, with leadership from the White House.
  • Shifting away from compliance toward a performance-based culture.
  • Adopting an enterprisewide approach to security rather than a system-based approach.
  • Developing a plan for collaborative research and development between government and industry.

Reader Comments

Wed, Mar 31, 2010 John

Well a couple of things. First let's look at that company who lost PII on 3.3 million student loan users from a lost USB stick. If someone had done C&A and then and actually asked if they encrypt any removable media then maybe they might have implemented that. Also, more important, Hmm, and who stands to make the most money if C&A goes away and we only focus on techies admining more and more IDS's...
Oh yeah, SANS... Hmmm.....

Wed, Mar 31, 2010 Dan Kuchem WDC

Providing IT security requires a layered solution. There is no one single point of failure. However, today's distributed computing systems have opened up security issues / holes that require enormous resources, but that in the end, also require vigilant end-user compliance to basic security awareness. One quickly evolving solution is to move toward a virtualized desktop environment, where data is maintained and controlled in the datacenter and the end-points can be locked down to control how data can be used, providing or restricting access to copy, print or store data locally. It essence, virtualized desktops provide the best of mainframe-type security and distributed computing access and convenience through any device with thin client access.

Wed, Mar 31, 2010 StrangeLoop

Security cannot be achieved through legislation; and the same problem FISMA had - wetware - will still be present with any new law. It will be interesting, however, to see how they manage to tuck in some means for the additional redistribution of wealth.

Wed, Mar 31, 2010 Owen Ambur Hilton Head Island, SC

Consistent with the open gov directive, it would be good if the plan for collaborative research and development between government and industry were posted on the Web in Strategy Markup Language (StratML) format (ANSI/AIIM 21:2009). The Federal Plan for Cyber Security and Information Assurance Research and Development is available in StratML format at http://xml.gov/stratml/drybridge/CSIAPlan.xml The goals established for the Information Systems Security LOB are available at http://xml.gov/stratml/drybridge/ISSLOB.xml ISPAB's plan is at http://xml.gov/stratml/drybridge/ISPAB.xml

Mon, Mar 29, 2010 InfoSec Pro Washington DC

Dr Paller has been ingenuous with his comments. FISMA is not and has not been a waste of money or time. Part of the problem is the required reporting from Congress. If it had adapted so it would show the level of security for sytems and not just the level of compliance, then FISMA would have worked better because resources would have been spent on security not on the paperwork. The documentation is definately needed, or else how will a new SysAdmin be able to ensure his system is protected if he has to start from scratch to evaluate the system to learn it security? The control testing is also needed, not just real-time monitoring. How will you defend your systems if you don't test the controls to know which ones are working and which are not? State has improved their security using FISMA and common sense. Agencies who are not willing to improve security are just completing the paper and nothing else. Agencies who want to improve their security need money, but they must make the case and if their FISMA score goes down, but the security goes up, it will prove the reporting is where the system breaks down.

Show All Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above