Tech apocalypse: Five doomsday scenarios for IT

Technology drives just about everything we do, and not just at our jobs. From banks to hospitals to the systems that keep the juice flowing to our homes, we are almost entirely dependent on tech. More and more of these systems are interconnected, and many of them are vulnerable. We see it almost every day.

But what if instead of simply a denial-of-service attack against select Websites, the entire Internet suddenly stopped working -- or for that matter, Google could not be reached. What if instead of a mere data breach, our financial institutions were attacked by a weapon that could instantly neutralize all electronic transactions? Or if hackers wormed their way into the systems that control the power grid?

[ Think you're prepared for every IT contingency? Take the InfoWorld worst-case-scenario quiz to find out. | Every bit of your virtual existence is being monitored -- get scared accordingly with our Top 10 reasons to be paranoid. ]

Heck, what if God decided she'd had enough of us and decided to send a solar storm our way?

If you think these things can't happen, think again. Some already have occurred on a smaller scale. But we thought it might be fun to turn up the volume and see what might happen -- how likely a "tech doomsday" scenario might be, how long it would take us to recover, and how we might prevent it from coming to be.

What could possibly go wrong? Try these scenarios for starters.

Tech doomsday scenario No. 1: America goes darkNews flash: A coordinated hack attack on our nation's power grid caused massive blackouts across the United States, leaving more than 300 million people without electricity for days.

The Supervisory Control and Data Acquisition (SCADA) systems that run U.S. power plants were built some 40 years ago, when the Internet was just a handful of university computers connected via 300-baud modems.

"Back then every power grid system in the world was considered its own island," says Robert Sills, CEO of RealTime Interactive Systems, which provides security solutions for industrial control applications. "There wasn't technology available to connect them. Now there is."

And the downside of all this connectivity is that once a local grid gets overloaded, others connected to it may tumble like dominoes. That's what happened in August 2003, when overgrown trees and human error triggered a power outage at Ohio's FirstEnergy. That failure caused a cascade that ultimately left 55 million people in the United States and Canada without power.

It doesn't take an act of God or Homer Simpson at the controls to cause a cascading power failure. It could be a rogue employee seeking revenge -- like the software engineer who hacked into an Australian water treatment plant's SCADA system in 1991, releasing 264,000 gallons of raw sewage.

Or it could be an external attacker who gains entry into a SCADA system's maintenance ports via war-dialing, and then uses social-engineering or spear-phishing attacks to gain entry into the network.

Sills says the vast majority of power substations are vulnerable to such an attack. From there, the attacker simply needs to change a few settings and let the grid's automated fail-safe systems do the rest.

"Right now it's a system that's pretty wide open," says Sills. "There are any number of ways someone could make unauthorized transactions via routine maintenance. You could create an outage simply by pushing the wrong key."

What could happen: Like the grid itself, other failures tend to cascade when the lights go out. In 2003, landline and cellular phone systems still worked but were so overloaded with calls that they effectively shut down. Electric railways stopped in their tracks, flights were canceled, and gas pumps would no longer pump. Water supplies that relied on electric filtering systems got contaminated. Food and medicine got spoiled; looting occurred; people died. On the positive side, residents of large cities were able to see the stars for possibly the first time in their lives.

How long would it take to recover: From hours to days, depending on how many generators have been affected and how long it takes to restart them, says Sills. Nuclear facilities can take several days, gas- and coal-fired generators require around 24 hours, but plants that use hydroelectric power may be able to get back online almost immediately. If an adjacent grid is still operating, the dark one may also be able to tap into its reserves.

Likelihood: Low. Electricity is supplied to the United States and Canada by eight separate, regional entities, so for the United States to go entirely dark would require a coordinated attack of key substations in each grid, says Sills. That makes a worldwide blackout even less likely. Still, regional blackouts are well within the grasp of knowledgeable attackers.

How to avoid this: The technology to secure the power grid is readily available. Sills says his firm has installed protective measures for a utility serving a major metro area, but declined to name it, lest it become a target. The problem? The Federal Energy Regulatory Commission is still hammering out security guidelines for the diverse systems used by power plants, and no public utilities are reluctant to invest in costly retrofits until their solution gets Uncle Sam's stamp of approval.

Tech doomsday scenario No. 2: Wall Street gets e-bombedNews flash: In what authorities suspect was the aftermath of an electromagnetic pulse weapon, a rogue attacker took down much of lower Manhattan today -- causing equipment failures and power outages on a massive scale and shutting down financial markets across the country.

Though most commonly associated with nuclear explosions, you don't need a nuke to create an electromagnetic pulse strong enough to do serious damage. EMP devices emit extremely high-frequency signals that fry electronics to a crisp, rendering them useless. An EMP will also wipe out or corrupt any data not stored on magnetic or optical devices. Worse, EMPs are largely untraceable, because the weapon itself destroys any evidence of its use.

A van with an EMP device in the back could effectively shut down big chunks of the U.S. economy simply by driving down Wall Street with the signal turned up, says Gale Nordling, CEO of Emprimus, a company that helps enterprises protect against threats from non-nuclear EMP.

If you wanted to take out the entire continent, though, you'd need a nuke and a missile delivery system. "One bomb exploded 300 miles over Kansas could take out most of the electronics in the United States," says Nordling.

What could happen: Workstations? Dead. Data centers? Gone. Cell phones might still work, but the cell towers probably won't, rendering them useless. Your car won't start. A large enough attack will also shut down automated controls at power substations, leaving everyone in the dark. Think pre-industrial revolution days. In our scenario the New York Stock Exchange shuts down, causing shock waves to reverberate throughout worldwide markets.

How long to recover: How long it takes organizations to bounce back depends on how serious they were about disaster recovery before hell broke loose. Backup power generators, fuel supplies, alternative work facilities, redundant data centers in multiple locations, and a well-rehearsed plan for making it all work together are the key elements to disaster recovery, says Richard Rees, security solution director for disaster recovery and business continuity specialists

Fortunately for our scenario, the financial sector is better prepared than most, says Rees.

"The best recent example are the financial institutions after 9/11," he says. "They had solid disaster recovery plans, they'd invested in their infrastructure and rigorously tested it, they knew what to do. They were back and open for business within three days. Their results were dramatically different than other organizations who'd tested their plans maybe once or twice. They could be out of commission for up to six months. There aren't too many businesses who can really withstand that."

Likelihood: Higher than you might think. You can buy a small EMP device over the Internet or download plans for building your own, says Nordling, who says he's been approached by a number of companies who believe they've already suffered an attack.

"There's a tremendous proliferation of information about EMP devices and the barriers to entry are extremely low," he says. "It's not just a tool for terrorists -- it could be disgruntled employees, criminals, extremists, competitors, or college kids who want to build one simply for the heck of it. From talking with members of Congress, they believe an EMP attack will happen. It's not a question of if, but when."

How to avoid this: One option is to install welded-steel shielding on all six sides of any room containing critical electronics, and put filters on all power and communications lines to siphon off high-frequency radio signals. A less costly option is to put your critical systems into a modular data center that's protected against EMP attacks, which you can fail over to when needed. Emprimus Director of Security Jim Danburg adds that some, but not all, Wall Street institutions are already protected.

Tech doomsday scenario No. 3: Google is goneNews flash: Visitors to Google.com were stunned when the world's dominant Web site returned a "404 Not Found" error for tens of millions of Web searchers. All Google services -- Gmail, Google Docs, AdSense - were inaccessible for periods ranging from hours to days, depending on users' locations.

Google has so insinuated itself into our lives it seems almost unthinkable that we might have to live without it. Experts consulted for this story agreed that to take down a company as mighty and well fortified would require someone on the inside -- not necessarily a malicious Google employee, just a stupid one (if such beings exist) with the right admin privileges.

It's not entirely unfeasible. Last December, attackers tricked Google employees to visit a malicious Web site, which then exploited a vulnerability inside Internet Explorer to install an encrypted backdoor into the Google network. From there they accessed the Gmail accounts of Chinese dissidents.

In our doomsday scenario, a Google employee merely installs a rogue application on the network that allows external attackers -- say, an unfriendly nation state with a grudge -- to slip behind the company firewall.

"The main vector for getting inside most organizations today are rogue applications residing on the network," says Nir Zuk, founder and CTO of Palo Alto Networks, a network security company.

For example: An IT manager installs GoToMyPC on a machine in the data center so that he can fix problems in the middle of the night from his home. But it has a weak password and gets hacked. Or he installs a P2P app to download songs, unwittingly allowing outsiders to download confidential files from the company LAN -- including password sets and network configuration maps. Or he sets up WebEx to do a presentation, then foolishly tells the program to share his desktop across the Web.

Once inside, attackers could root around the network until they locate the command and control centers for Google's many data centers. And then they can turn out the lights, leave behind a logic bomb that corrupts Google's databases, or simply have their way.

"I'm not familiar with the structure of Google's network, but they must have a command and control app that lets them shut down their data centers," says Zuk. "Everyone does."

What could happen: Yahoo and Bing become swamped with search traffic, and might collapse under the weight. Organizations that rely on Gmail and Google Docs for their day-to-day operations will find themselves unable to get much done (though, given how many outages Gmail had over the last year, they might be used to it). YouTube fans may discover there are approximately 7,834 other free video sites out there. Web entrepreneurs who rely on Google ads will find themselves bereft of income for an unknown period of time.

Other consequences, according to Google Blogoscoped author Philipp Lenssen: "People may not be able to post an update about their life, leading others to believe they've disappeared (because Blogspot is down); conspiracy theorists will be able to sell more books on 'why Google went down (and what the NSA had to do with it)'; and people who want to search for 'why Google is down' realize that, well, Google is down so they can't search for that."

How long it would take to recover: From hours to days, depending on what measures Google already has in place. A Google spokesperson contacted for this story says, "We are always planning for different threat scenarios, but we aren't going to discuss specific defense measures."

Likelihood: Zuk says it's more likely than most big companies are willing to admit.

"In a big company like Google or Yahoo, which have tens of thousands of employees, there will always be unaware employees who do something stupid like sharing their desktop via WebEx," he says. "It only takes one to do it, and from there the route to the data center is a quick one."

How to avoid this fate: To avoid getting nailed by rogue apps, companies need greater visibility into their networks to expose any apps that are running and what ports they are using, and to map all of their other dependencies as well, says Steve Cotton, CEO of FireScope, a developer of IT service management solutions.

1 2 Page 1
Computerworld's IT Salary Survey 2017 results
Shop Tech Products at Amazon