While the prospect of damaging infrastructure through computer worms is now a real threat, IBM accumulates a small arsenal of technologies to fight the inevitable battle to protect our smarter, but more fragile, planet.
It's quite common for geeks to be science fiction fans, and I don't know if I'm a geek, but I do like a good sci-fi movie. Thus, I can't wait until tonight to go see Battle: Los Angeles, which opens today and tells the story of a Marine platoon's efforts to repel an alien invasion of the city of angels. Science fiction battles, like actual war, are glorified in the abstract, but you wouldn't want to be there if it were happening in real life. Thus it is with grave concern that I read about the cyber attacks on Iranian nuclear facilities using what is believed to be the Stuxnet worm.
Speculation is that Israel, with the support of the United States, tested the effect of the Stuxnet worm at a secret facility in Israel's Negev desert. Agents reportedly were spinning nuclear centrifuges similar to those located in Iran at Dimona, an Israeli nuclear arms complex, according to the New York Times. Apparently Iran's nuclear enrichment program, which uses the centrifuges, as well as a nuclear power plant under construction at Bushehr, were damaged by the worm. There is debate over the extent of the damage, and the Iranians deny that the worm caused any significant setbacks, though they concede it was found on technicians' laptops. Others, however, say the nuclear reactor was at risk of an actual meltdown had it been started up following the cyber attack.
While some are saying that disabling Iran's nuclear facilities may offset or delay military action against the Mideast country with persistent nuclear ambitions, the countermeasure may go down in history as one of the first examples of cyberwarfare or cyberterrorism. To understand what is occurring, you have to realize that a lot of machines and physical equipment in today's complex societies are run by computers. The equipment is controlled by what are known as SCADA systems, which stands for "supervisory control and data acquisition." These systems monitor and control physical processes, including things like traffic lights as well as some biggies like electricity transmission, oil and gas pipelines, and water distribution. Oh—and let's add nuclear power plant operation to that list.
Of course, all the important SCADA systems in North America and Europe are hardened against any sort of attack, from computer worm and other threats, right? Well, we wish that were the case. The truth is, SCADA-based systems are quite vulnerable to cyberwarefare attacks, and some people are quietly waiting for the other shoe to drop. What do they say about people living in glass houses?
Stuxnet was first reported on publicly in June 2010, but its presence probably goes back to the prior year. It appears to target Siemens' WinCC/PCS7 systems running on MS Windows. It uses four zero-day attacks to install a rootkit. This logs into the SCADA database and "steals design and control files," according to one report. Apparently, the program can change the control system and then even hide its changes.
Such a worm has the ability to cause untold damage to today's industrial, infrastructure, or facility-based processes. Imagine if it got into the systems controlling the electricity grid and caused unpredictable and hard-to-trace power outages. Most water today has to be treated before people can drink it. If the water treatment plants all were to shut down at the same time, how much potable water is on hand, and how long would it last? It's difficult to imagine what might happen if the radar and communications suddenly went haywire at a major metropolitan airport. Such occurrences could result if the systems became infected by Stuxnet or a similar worm.
Despite the potential severity of such an attack, there is a general lack of concern about security in SCADA-based systems. Some believe they are protected by their relative obscurity and proprietary interfaces; others mistakenly believe they are safe because the systems may not be connected to the Internet. Both assumptions are almost dreamlike in their naiveté. The truth is, SCADA systems are vulnerable to unauthorized access to the control software, either through a virus or human intervention, as well as access to the networks connecting SCADA devices. While protection of the network with cryptographic devices is assumed to block any access to the SCADA control software, the network jacks and switches provide a means for bypassing security. And in the case of public agencies charged with protecting many of these critical systems, there isn't a lot of extra money floating around today to spend on beefed-up security.
The security industry and independent organizations are well aware of the national threat and are quietly making inroads to at least to create some standards against which companies and agencies may be evaluated, but progress is slow. The hope is that they will complete their work before "something happens."
IBM also is well aware of the increasing threat to SCADA, as well as other computer systems running operations of various large public and privately held corporations. The company has nine worldwide research labs innovating security technology and nine security operations centers around the world to support clients. It has acquired no fewer than four security firms in about a year—Watchfire, Ounce Labs, Telelogic, and BigFix, many of which focus on smart devices, themselves raising untold concerns among IT security professionals.
IBM Security Solutions now has an extensive portfolio of hardware, software solutions, and professional and managed services that address the full range of IT security risks. It recently announced a collaboration with Trend Micro in which it integrates the company's ant-malware solution into IBM's new Tivoli Endpoint Management platform. The company also just introduced a new network security appliance, the IBM Security Network Intrusion Prevention System (IPS) GX7800, which protects Web applications and data in the cloud, all while operating at an estimated 20 GBps.
"The endpoint of 2011 is no longer just a PC or laptop," says Steve Robinson, general manager of IBM Security Solutions. It's "the entire range of interconnected and instrumented devices and sensors that comprise the smarter planet," he says. "From electrical grids to mobile devices, transportation systems, and buildings, the proliferation of these intelligent systems is creating new security loopholes that businesses need to address," says Robinson. "On today's smarter planet, everything is an endpoint and must be managed and secured like any other critical business asset."
By 2015, it is expected there will be 300 million smart electrical meters deployed worldwide. While each has the potential to save money and improve efficiency, they can also present an opportunity for a security breach, since they can sense and interact with other devices.
Protecting all these computer systems that constitute the smart planet isn't going to come cheap. As the risks associated with securing these systems and endpoints becomes increasingly complex, the cost of managing and securing them will rise. IDC estimates that the endpoint security market will increase at a compound annual growth rate of 8.3 percent to reach nearly $10 billion by 2014.
The race against time appears to be on, and one wonders what form the security breaches will take before a major SCADA system—outside of Iran—is infected or damaged. Will all the companies and public agencies that need to protect their systems be able to do so, and have the money to act, before "something happens"?
I suppose if you're in the security industry, your work is cut out for you. You might even declare: "Let the cyberwars begin!" Me, I'll be satisfied with a new interpretation of a good old alien invasion—the kind that's not real.