With major cyber-security initiatives by the Department of Homeland Security underway, and the U.S. House of Representatives passing nearly $400 million in IT security research, I wonder if the efforts are being placed where they are most needed, and if more would be achieved by focusing on application security - and unleashing the bug finders.Testifying in front of the Senate Intelligence Committee earlier this week, Dennis C. Blair, Director of National Intelligence warned that U.S. critical infrastructure is "severely threatened" and called the recent cyber attack on Google "a wake-up call to those who have not taken this problem seriously."
From Thomas Claburn's story U.S. 'Severely Threatened' By Cyber Attacks:
"Sensitive information is stolen daily from both government and private sector networks, undermining confidence in our information systems, and in the very information these systems were intended to convey," said Blair in prepared remarks outlining the U.S. intelligence community's annual assessment of threats.
Blair also explained how cyber-criminals' capabilities exceed the ability of companies and agency's ability to respond. He called for more effective public and private sector IT security collaboration.
Also last week, DHS is looking to invest nearly $900 million in fiscal 2011 on technology projects that include bolstering cyber security and consolidating its data centers.
According to Elizabeth Montalbano's story Homeland Security Plans Cybersecurity, Data Center Investments:
DHS is asking for $379 million to go to its National Cyber Security Division (NCSD) to develop capabilities for preventing and responding to cyber attacks. The department plans to use the money to identify and reduce vulnerabilities within both its .gov and .com Internet domains, officials said on a conference call.
NCSD is a division within DHS that's meant to work collaboratively with public, private, and international organizations to secure cyberspace and the U.S. government's cyber infrastructure.
Increased spending on IT security research, awareness, and education are all great things, and we've needed more investment in these areas for 20 years now. Improved public and private IT security collaboration and data sharing are well worth the effort, too: especially if that data sharing is a two-way street, with data flowing from the government to the private sector and not just the other way around.
Still, I don't get a sense that we are focusing on the top priority when it comes to cyber security: the abysmal security level that most applications ship. Most software today comes loaded with security holes. This week, Microsoft is publishing 13 security bulletins, but it's not yet fixing a flaw revealed in Internet Explorer last week.
And the recent attacks on Google and dozens of other western companies were made too easy by flaws in Internet Explorer.
Microsoft is not the singular source of these problems. In fact, Microsoft has made significant improvements since its Trustworthy Computing initiative. It's a systemic problem throughout the software industry. Pick a major software maker - any one - and you are going to find security flaws a Navy armada could pass through.
I've been harping about software quality, as it applies to security, for nearly a decade - and until we get a handle on it there isn't going to be any form of "cyber-security." You can try to protect yourself all you want, but if the very software in use makes it possible for an attacker to need only one employee to open one maliciously-crafted attachment or link to infiltrate a company - we have already lost.
I don't know what the answer will eventually turn out to be. Perhaps if more software vendors and organizations that develop software in-house did more of what Herbert H. Thompson describes in his article Secure Software Needs Careful Testing--And Lots Of It. Or, maybe if companies simply refused to run insecure software, as Alexander Wolfe proffers in Sloppy Software Dev Exposes Google Hacker Holes vendors would notice and do more about the problem they're creating.
Imagine if the Federal government and many Fortune 500 companies started ditching the applications that enable them to be attacked, and refused to buy any software until it reached a certain level of trust and security? That would get the software industry's attention. The problem is that there wouldn't be much software left to procure.
Maybe the solution is somewhere within idea of software liability. The problem with that idea, that has always concerned me, is the impact it would have on software innovation and cost.
Or, maybe it's time we provide incentives for the software security researchers to do what they do best: find bugs.
Unleash the Bug Finders Perhaps the Federal government could hire a few thousand ethical hackers and application security experts, and offer them legal cover, to go find all of the software flaws they're able to find. Then require software companies to rapidly fix them. Such a program could evolve into an ongoing bounty the federal government would pay the bug finders.
The program could even be paid for by a tax levied on the software industry. Or, perhaps, the software vendor is fined for each flaw, and that fine pays the bounty. As it stands now, it's the software companies customers that pay the tax in the form of unending patch updates and attacks on their systems. Also, as it stands now, software vendors don't pay high enough a price for the flaws they inflict.
The bounty would also provide competition to the underground market for these software flaws. The underground software bug market is funded by organized crime and state-sponsored hackers. It could use more competition than a handful of security vendors who pay for bugs.
Whatever the solution eventually turns out to be, it's going to require an Apollo 11 level of effort. And it will require commitment by government and the buyers of software to force software companies to stop publishing insecure code.
And it's time to put more ideas on the table. And we should be open to consider anything, as the status quo of software quality can't stand as it is.