The pressure is on President-elect Obama to make cyber-security a priority issue. What can the government, let alone the president, do about this? Obama has said, “As president, I’ll make cyber-security the top priority that it should be in the 21st century. I’ll declare our cyber-infrastructure a strategic asset, and appoint a national cyber-adviser, who will report directly to me.”
My guess is that President Obama’s national cyber-adviser will run into the same problem that previous attempts have found: No section of the federal government is interested in giving up control over the security of its own computers. There can be, and are, standards for the security of systems in federal networks. The standards are modest and not well adhered to, and they don’t include any clear penalty for noncompliance.
If the implication of Obama’s plan is that there will be some IT security czar or agency in charge-currently the OMB (Office of Management and Budget) seems to have the main role in the standards I mentioned-of setting and enforcing IT security rules, I have to say it’s hard to imagine him, her or it succeeding.
There were a bunch of people more or less in this role in the Bush administration. Remember Richard Clarke? Amit Yoran? Greg Garcia? Some people would dismiss these people as … well, they worked for Bush so they must be corrupt or incompetent or, in Richard Clarke’s case, defeated by the forces of corruption and incompetence. I figure the truth is different and more discouraging.
The Real Challenge to Government Cyber-security
A long time ago I worked for a company that did some contract testing work for the Canadian federal government. (I spent a February in Ottawa and Hull. Brr …) The government officials we dealt with liked to point out that all Canadian computer purchasing was centralized, but in the United States it was distributed to each department. They thought this was nuts and horribly wasteful and they had a point. This alone makes centralizing security management difficult, if at all possible.
But the same is true of the management of security, which is intrinsically tied in with other policies of computer and network usage and access. Security policy dictates what systems software you can use, what your update policies are, what your policies must be for remote access to computers and so on. The justification for letting individual departments, or even smaller units, make these decisions is that they understand their own needs better than some Federal Office of Security Autocracy.
These departments and agencies don’t want to surrender authority, let alone budget dollars, to someone outside who will be telling them what they can do with their own computers. It’s a serious loss of power and affects their policy missions. Call it a turf war if you want and say what it needs is a good dose of leadership, but that’s basically just rhetoric. What we’re talking about here is a major, radical change, and it’s not likely.
The Defense Science Board report (PDF) quoted in eWEEK seems to be arguing in part for “deperimeterization,” an idea I’ve ridiculed in the past.. But even being generous to the idea, it demands even more fundamental, radical change and even more sophisticated management tools. It’s not realistic. The only changes we should expect to see are incremental.
So, yeah, our cyber-infrastructure is a strategic asset and protecting it is a priority. This is not a bold, controversial position. Some experts may argue that cyber-security wasn’t a priority for the Bush administration, but I’ll be surprised if it ends up being much more of one for the next administration.
Security CenterEditor Larry Seltzer has worked in and written about the computer industry since 1983.
For insights on security coverage around the Web, take a look at eWEEK.com Security Center Editor Larry Seltzer’s blog Cheap Hack.