Security training is a waste of time and money.
That declaration from Immunity Inc. CEO Dave Aitel last month produced a predictable range of responses from security experts.
A few of his fellow security experts say he's right; some say he's crazy. Many think he raised valid points, but is attacking a straw man -- that nobody in the industry thinks training is foolproof or a guarantee of security, but is still key to "layered" security. Calling security training worthless, they argue, is like calling driver training worthless if young drivers get pulled over for violations or have accidents.
Was Aitel, a former computer scientist for the National Security Agency, being intentionally provocative?
No, he said in an interview Wednesday. "The truth is usually provocative. I'm not one of those people who is going to tell people what they already know."
Aitel said he has no statistical evidence for his claim, but noted that security firms are stingy about sharing data on how effective training is. And anecdotally, he said it's clear that even trained employees don't have the savvy to defeat sophisticated attacks.
Instead, he argued that IT professionals who seek training programs for users are trying to duck their own responsibility.
Among those who have weighed in on Aitel's views is Mark Baldwin, principal researcher and consultant at InfosecStuff. He argued that Aitel "paints it as an either-or situation. In fact, security awareness is part of a strategy of defense in depth. Of course, businesses should secure and monitor their networks, isolate and protect their data and have a well-defined incident response plan. But they should also have a robust security awareness program."
A number of online comments in response to Aitel offered the same view. "I do conduct frequent information security awareness and competence audits for organizations and I have seen the security risks considerably reduced due to better employee awareness and security skills," said AnupNarayanan.
That sentiment was also echoed on the Sophos Naked Security blog, which posted an entry on the topic.
One Sophos blog reader, "authorizedpants" wrote: "Of course some attacks will get through. That's why my department is called IT Risk *Management* and not IT Risk Elimination. The point is to have defense in depth and that includes educated end users so they become less likely to click on things they shouldn't. It also includes AV and proxies and IDS and firewalls and monitoring."
Jody Westby, CEO of Global Cyber Risk and a consultant on privacy, security, cybercrime and IT governance, said the flaw in Aitel's argument is that, "with everyone using their own devices and social media, training has become all the more important." The failure to train employees, she added, could violate compliance requirements and expose enterprises to legal liability.
Aitel remains unrepentant. "I don't' believe (security training) has a role," he said. "I'm not taking it back, because it happens to be correct. Even if you're reducing your click-through rate to 5 percent, you haven't accomplished anything, because the attacker is inside."
Some experts like Kevin McAleavey, cofounder and chief architect of the KNOS secure desktop project, feel that Aitel is correct.
The vulnerability of end users, he said, is the main reason for the development of KNOS software. "We realized then that the only way to protect a network was to lock down the end-user desktops so that whatever they did couldn't harm them or the network. If malware is unable to gain a foothold, then it does nothing."
McAleavey notes that cybercriminals know what kind of training employees get and quickly find ways to work around it. When employees were trained to open emails only from those they trusted, "what did the malware folks do? They hit the servers, got the email addresses of people that the end user knew and faked the emails with malware attachments from their 'friends' lists."
Security programs, he said, "should have picked off that email before it ever landed in the employee's inbox."
McAleavey isn't entirely anti-training. "It does help," he said. "It leaves only the hardcore stupid to infect the network. But the better solution is to keep them from doing any harm in the first place."
Aitel even questions those who say the problem is ineffective training, and that "effective" training has value.
"Effective at what is a good question," he said. "It's not effective against targeted social engineering. There are cheaper, easier technological methods (to provide security). I'd much rather have web proxy than training programs, which are fairly expensive."