Raytheon's cyberchief describes 'Come to Jesus' moment

After Raytheon began selling missiles to Taiwan in 2006, the defense company's computer network came under a torrent of cyberattacks.

"We truly had the 'come to Jesus moment' five years ago because we decided ... to sell missiles to Taiwan," said Vincent Blake, head of cyber security at Raytheon U.K., during a panel session at the RSA security conference in London on Wednesday.

"For some reason, a country next door to Taiwan didn't really like that so they got very interested in our IPR [intellectual property rights]," he said. "We've had to very, very rapidly catch up with our own internal networks."

Blake described a "huge leap in attacks" that prompted the company to make cybersecurity one of its top five priorities, and eye security companies for acquisition. Since that time, Raytheon has continued to be an attractive target for hackers, given its breadth of defense technologies that supply militaries around the world.

Now, the company sees an incredible 1.2 billion -- that's billion -- attacks on its network per day, Blake said. About 4 million spam messages target Raytheon's users, and the company sees some 30,000 samples per day of so-called Advanced Persistent Threats, or stealthy malware that seeks to stay long-term on infected computers and slowly withdraw sensitive information.

"We are the most targeted industry in the world," Blake said.

So how does Raytheon defend itself?

Raytheon uses sophisticated analysis engines that can sort through network alerts, Blake said. Some decisions are automated, while other alerts are assigned to a dedicated analyst for investigation.

Zero-day exploits, or attacks actively being used on the Internet against vulnerabilities that do not have a patch, are a big problem, said Blake, speaking to the IDG News Service after the panel. Last year, Raytheon detected 138 zero-day attacks against some 5,000 employees, he said.

The zero-day attacks were detected through RShield, a Raytheon product that examines e-mail attachments and embedded URLs. If an e-mail attachment comes through Raytheon's system, it is first scanned through commercial antivirus software and then through RShield, which scans the attachment in a hypervisor, Blake said.

The hypervisor is custom-built and not VMware, Blake said. Many hackers engineer their malware to not execute within VMware. The behavior of the attachment is observed, and if it does something suspicious, it is blocked. Blake said it's the only way these days to detect advanced malware.

"That's where the future is," Blake said. "If you haven't seen it [the malware] before, you're not going to find it."

Last week, Blake said Raytheon saw its first cloud-based attack on its network: 20 Raytheon employees received a targeted e-mail with a link to an application hosted with a cloud service provider. The style of attack -- a malicious email -- is a typical social engineering technique known as spear phishing that can give hackers an easy foothold in an organization. Unfortunately, two people clicked on the link, Blake said.

Blake said his team was able to detect the attack once the affected Raytheon computers started "beaconing" to the cloud service provider, or trying to make a network connection.

Raytheon only lets an attacker sit on its network for two hours or less, a response time that Blake said the company hope to cut down to 10 minutes.

"You will be attacked," Blake said. "You will be exploited. It's not a matter of whether something will get in your system, but more how long you will continue to have them in your system."

In March, Blake said Raytheon mounted a "companywide response" when RSA revealed that part of its SecurID system had been compromised in March. Passwords were changed. Raytheon still uses SecurID but has since added other layers of security in their authentication systems.

Due to the breach, "we had to significantly change our attitude to not being so reliant on RSA," he said.

Send news tips and comments to jeremy_kirk@idg.com