The attackers chose their moment well.
On Apr. 7, 2011, five days before Microsoft patched a critical zero-day vulnerability in Internet Explorer that had been publicly disclosed three months earlier on a security mailing list, unknown attackers launched a spear-phishing attack against workers at the Oak Ridge National Laboratory in Tennessee.
The lab, which is funded by the U.S. Department of Energy, conducts classified and unclassified energy and national security work for the federal government.
The e-mail, purporting to come from the lab's human resources department, went to about 530 workers, or 11 percent of the lab's workforce.
The cleverly crafted missive included a link to a malicious webpage, where workers could get information about employee benefits. But instead of getting facts about a health plan or retirement fund, workers who visited the site using Internet Explorer got bit with malicious code that downloaded silently to their machines.
Although the lab detected the spear-phishing attack soon after it began, administrators weren't quick enough to stop 57 workers from clicking on the malicious link. Luckily, only two employee machines were infected with the code. But that was enough for the intruders to get onto the lab's network and begin siphoning data. Four days after the e-mails arrived, administrators spotted suspicious traffic leaving a server.
Only a few megabytes of stolen data got out, but other servers soon lit up with malicious activity. So administrators took the drastic step of severing all the lab's computers from the internet while they investigated.
Oak Ridge had become the newest member of a club to which no one wants to belong – a nonexclusive society that includes Fortune 500 companies protecting invaluable intellectual property, law firms managing sensitive litigation and top security firms that everyone expected should have been shielded from such incursions. Even His Holiness the Dalai Lama has been the victim of an attack.
Last year, antivirus firm McAfee identified some 70 targets of an espionage hack dubbed Operation Shady RAT that hit defense contractors, government agencies and others in multiple countries. The intruders had source code, national secrets and legal contracts in their sights.
Source code and other intellectual property was also the target of hackers who breached Google and 33 other firms in 2010. In a separate attack, online spies siphoned secrets for the Pentagon's £185 billion Joint Strike Fighter project.
Then, last year, the myth of computer security was struck a fatal blow when intruders breached RSA Security, one of the world's leading security companies that also hosts the annual RSA security conference, an august and massive confab for security vendors. The hackers stole data related to the company's SecurID two-factor authentication systems, RSA's flagship product that is used by millions of corporate and government workers to securely log into their computers.
Fortunately, the theft proved to be less effective for breaking into other systems than the intruders probably hoped, but the intrusion underscored the fact that even the keepers of the keys cannot keep attackers out.
Independent security researcher Dan Kaminsky says he's glad the security bubble has finally burst and that people are realising that no network is immune from attack. That, he says, means the security industry and its customers can finally face the uncomfortable fact that what they've been doing for years isn't working.
"There's been a deep conservatism around, ‘Do what everyone else is doing, whether or not it works.' It's not about surviving, it's about claiming you did due diligence," Kaminsky says. "That's good if you're trying to keep a job. It's bad if you're trying to solve a technical problem.
In reality, Kaminsky says, "No one knows how to make a secure network right now. There's no obvious answer that we're just not doing because we're lazy."
Simply installing firewalls and intrusion detection systems and keeping anti-virus signatures up to date won't cut it anymore - especially since most companies never know they've been hit until someone outside the firm tells them.
"If someone walks up to you on the street and hits you with a lead pipe, you know you were hit in the head with a lead pipe," Kaminsky says. "Computer security has none of that knowing you were hit in the head with a lead pipe."
According to Richard Bejtlich, chief security officer for computer security firm Mandiant, which has helped Google and many other companies conduct forensics and clean up their networks after an attack, the average cyberespionage attack goes on for 416 days, well over a year, before a company discovers it's been hacked. That's actually an improvement over a few years ago, he says, when it was normal to find attackers had been in a network two or three years before being discovered.
Bejtlich credits the drop in time not to companies doing better internal monitoring, but to notifications by the FBI, the Naval Criminal Investigative Service and the Air Force Office of Special Investigation, who discover breaches through a range of tactics including hanging out in hacker forums and turning hackers into confidential informants, as well as other tactics they decline to discuss publicly. These government agencies then notify companies that they've been hacked before they know it themselves.
Shawn Henry, the FBI's former top cyber-cop, is gravely warning that corporate hacking is much worse than people think it is.
But even the FBI took a defeatist view of the situation recently when Shawn Henry, former executive assistant director of the FBI, told The Wall Street Journal on the eve of his retirement from the Bureau that intruders were winning the hacker wars, and network defenders were simply outgunned.
The current approaches to fending off hackers are "unsustainable," Henry said, and computer criminals are too wily and skilled to be stopped.
So if hackers are everywhere and everyone has been hacked, what's a company to do?
Kaminsky says the advantage of the new state of affairs is that it opens the window for innovation. "The status quo is unacceptable. What do we do now? How do we change things? There really is room for innovation in defensive security. It's not just the hackers that get to have all the fun."
Companies and researchers are exploring ideas for addressing the problem, but until new solutions are found for defending against attacks, Henry and other experts say that learning to live with the threat, rather than trying to eradicate it, is the new normal. Just detecting attacks and mitigating against them is the best that many companies can hope to do.
"I don't think we can win the battle," Henry told Wired.com. "I think it's going to be a constant battle, and it's something we're going to be in for a long time…. We have to manage the way we assess the risk and we have to change the way we do business on the network. That's going to be a fundamental change that we've got to make in order for people to be better secure."
In most cases, the hacker will be a pedestrian intruder who is simply looking to harvest usernames and passwords, steal banking credentials or hijack computers for a botnet to send spam.
These attackers can be easier to root out than focused adversaries - nation states, economic competitors and others - who are looking to steal intellectual property or maintain a strategic foothold in a network for later use, such as to conduct sabotage in conjunction with a military strike or in some other kind of political operation.
Once a company's networks have been breached, Bejtlich says his company focuses on finding all of the systems and credentials that have been compromised and getting rid of any backdoors the intruders have planted. But once the attackers have been kicked off the network, there is generally a flood of new attempts to get back into the network, often through a huge wave of phishing attacks.
"For the most part, once you've been targeted by these guys, you're now living with this for the rest of your security career," Bejtlich said.
Many companies have resolved themselves to the fact that they're never going to keep spies out entirely of their network and have simply learned to live with the intruders by taking steps to segregate and secure important data and controls.
Henry, who is now president of CrowdStrike Services, a newly launched security firm, says that once companies accept that they're never going to be able to keep intruders out for good, the next step is to determine how they can limit the damage. This comes down, in part, to realising that "there are certain pieces of information that just don't need to reside on the network."
"It comes down to balancing the risks, and companies need to assess how important is it for me to secure the data versus how important is it to continue doing my business or to be effective in my business," he says. "We have to assume that the adversary is on the network and if we assume that they're on the network, then that should change the way we decide what we put on the network and how we transmit it. Do we transmit it in the clear, do we transmit it encrypted, do we keep it resident on the network, do we move it off the network?"
Bejtlich says that in addition to moving data off the network, the companies that have been most successful at dealing with intruders have redefined what's trustworthy on their network and become vigilant about monitoring. He says there are some organisations who have been plagued by intruders for eight or nine years who have learned to live with them by investing in good detection systems.
Other companies burn down their entire infrastructure and start from scratch, going dark for a week or so while they re-build their network, using virtualisation tools that allow workers to conduct business while protecting the network core from attackers.
Bejtlich, who used to work for General Electric, said one of the first things he did after being hired by GE was to establish a segmented network for his security operations, so that any intruders who might have already been on the corporate network wouldn't have access to his security plans and other blueprints he developed for defending the network.
"The first thing you've got to do is to establish something that you trust because nobody else can get access to it, and then you monitor the heck out of it to see if anybody else is trying to poke around," he said. "So you go from a posture of putting up a bunch of tools and sitting back, to one of being very vigilant and hunting for the bad guys…. The goal is to find them so quickly that before they can really do anything to you to steal your data, you've kicked them out again."
Kaminsky advocates shrinking perimeters to limit damage.
"Rather than one large server farm, you want to create small islands, as small as is operationally feasible," he says. "When you shrink your perimeter you need to interact with people outside your perimeter and figure out how to do that securely" using encryption and authentication between systems that once communicated freely.
"It changes the rules of the game," he says. "You can't trust that your developers' machines aren't compromised. You can't trust that your support machines aren't compromised."
He acknowledges, however, that this is an expensive solution and one that not everyone will be able to adopt.
While all of these solutions are more work than simply making certain that every Windows system on a network has the latest patch, there's at least some comfort in knowing that having a hacker in your network doesn't have to mean it's game over.
"There have been organisations that this has been like an eight- or nine-year problem," Bejtlich says. "They're still in business. You don't see their names in the newspaper all the time [for being hacked], and they've learned to live with it and to have incident detection response as a continuous business process."
Kim Zetter is a senior reporter at Wired covering cybercrime, privacy, security and civil liberties.