Stepan Gorsky: This violates the Geneva convention.
Jed Eckert: I never heard of it!
But knowing there is a growing threat against us from cyber attack, the issues raised are very legitimate, From this month’s Foreign Policy Magazine.
In Cyberwar, There are No RulesDon't agree with all the points, but a very thought provoking article and it's worth the full read.
Why the world desperately needs digital Geneva Conventions.
Tarah Wheeler September 12, 2018
In 1984, a science fiction movie starring an up-and-coming Austrian-American actor took the box office by storm. A cybernetic organism is sent back in time to seek out and kill the mother of a great war hero to prevent his subsequent birth. The cyborg scans a phone book page and begins methodically killing all women named Sarah Connor in the Los Angeles area, starting at the top of the list.
If The Terminator were set in today’s world, the movie would have ended after four and a half minutes. The correct Sarah Connor would have been identified with nothing but a last name and a zip code—information leaked last year in the massive Equifax data breach. The war against the machines would have been over before it started, and no one would have ever noticed. The most frightening thing about cyberwarfare is just how specifically targeted it can be: An enemy can leap national boundaries to strike at a single person, a class of people, or a geographic area.
Nor would a cyborg be necessary today. According to U.S. census data, there are currently 87 people in the United States named Sarah Connor. Many of them probably drive cellular-enabled cars that run outdated firmware, use public unencrypted Wi-Fi, and visit doctors who keep unsecured health care records about patient allergies and current medications on computers running the infamously outdated and vulnerable Windows XP operating system.
So far, the U.S. government has fumbled on cybersecurity, outsourcing much of that area of conflict to the private sector in accordance with the Trump administration’s most recent National Security Strategy—leaving the country exposed to foreign attack.
These days, warfare is conducted on land, by sea, in the air, across space, and now in the fifth battleground: cyberspace. Yet so far, the U.S. government has fumbled on cybersecurity, outsourcing much of that area of conflict to the private sector in accordance with the Trump administration’s most recent National Security Strategy—leaving the country exposed to foreign attack.
Those third parties operate under exactly the same incentives as any pharmaceutical company. If a company’s service is the treatment of symptoms, preventive medicine is a threat to its business model. Meanwhile, pundits, policymakers, and publishers take as gospel what they’re told by so-called cybersecurity experts who have more social media followers than relevant credentials in the field, which is how hysterical “The Hackers Are Coming for Us” editorials find their way into otherwise respectable publications.
Increased fear, uncertainty, and doubt surrounding cybersecurity have led to a world where we cannot tell what has and hasn’t happened. The nature of cyberwarfare is that it is asymmetric. Single combatants can find and exploit small holes in the massive defenses of countries and country-sized companies. It won’t be cutting-edge cyberattacks that cause the much-feared cyber-Pearl Harbor in the United States or elsewhere. Instead, it will likely be mundane strikes against industrial control systems, transportation networks, and health care providers—because their infrastructure is out of date, poorly maintained, ill-understood, and often unpatchable. Worse will be the invisible manipulation of public opinion and election outcomes using digital tools such as targeted advertising and deep fakes—recordings and videos that can realistically be made via artificial intelligence to sound like any world leader.
The great challenge for military and cybersecurity professionals is that incoming attacks are not predictable, and current strategies for prevention tend to share the flawed assumption that the rules of conventional war extend to cyberspace as well. Cyberwarfare does have rules, but they’re not the ones we’re used to—and a sense of fair play isn’t one of them. Moreover, these rules are not intuitive to generals versed in fighting conventional wars.
That’s a problem because cyberwar won’t be waged with the informed participation of much of the U.S. technology sector, as the recent revolts at Google over AI contracts with the U.S. Defense Department and at Microsoft over office software contracts with U.S. Immigration and Customs Enforcement demonstrate. That leaves only governments and properly incentivized multinational corporations to set the rules. Neither has yet provided a workable and operational definition of what constitutes a globally recognized act of war—a vital first step in seeking to prevent such transgressions.
The closest that the U.S. military has come to such a definition is to say that “acts of significant consequence” would be examined on a case-by-case basis and could require congressional evaluation. But given how quickly a cyberattack could disable critical infrastructure, expecting Congress to react in time to answer effectively is unrealistic.
In a world where partisan politics have been weaponized, a smart misinformation campaign by a foreign state that targeted only one political party might even be welcomed by other parties so long as there was plausible deniability—and with cyberattacks, attribution is rarely certain.
There is also a serious risk of collateral damage in cyberoperations. Most militaries understand that they are responsible not only for targeting strikes so that they hit valid targets but also for civilian casualties caused by their actions. Though significant collateral damage assessment occurs prior to the United States authorizing cyberoperations, there is no international agreement requiring other powers to take the same care.
A major cyberattack against the United States in 2014 was a clear example of how civilians can bear the brunt of such operations. Almost all cybersecurity experts and the FBI believe that the Sony Pictures hack that year originated in North Korea. A hostile country hit a U.S. civilian target with the intention of destabilizing a major corporation, and it succeeded. Sony’s estimated cleanup costs were more than $100 million. The conventional warfare equivalent might look like the physical destruction of a Texas oil field or an Appalachian coal mine. If such a valuable civilian resource had been intentionally destroyed by a foreign adversary, it would be considered an act of war.
In the near future, attacks like the Sony hack will not be exceptional. There are countless vulnerabilities that could result in mass casualties, and there are no agreed norms or rules to define or punish such crimes. Consider the following examples.
The United States Is Not Ready for a Cyber-Pearl Harbor
The weekend’s massive “ransomware” attack exposed the glaring vulnerabilities in our cybersecurity readiness.
Once a week, a European aircraft manufacturer cleans all plane cockpits of Android malware. Pilots can pass malware to the plane from their smartphones when they plug them in, which the plane (while theoretically unaffected by phone-only malware) then passes it on to the next pilot with a smartphone. Planes are already covered in viruses, both virtual and microbial. In such a vulnerable environment, even an unsophisticated hack could wreak havoc. A text message sent to the phone of every in-air pilot giving them a national security warning or rerouting their planes could lead to emergency landings and widespread confusion, with more sophisticated attacks potentially leading to far more serious consequences.
Aviation is not the only vulnerable sector. The U.S. health care system is full of medical devices running ancient firmware or operating systems that simply cannot be patched or hardened against commonly known network intrusions. Small hospitals often cannot afford to replace their medical equipment on a regular schedule, and device providers may deprioritize or block security patches or upgrades in order to sell updated devices in the next round of production.
That’s a problem in an era when many surgical procedures are assisted by robots, which hospitals struggle to keep secure. The medical device industry focuses more on performance and patient health outcomes than on keeping a cyberadversary at bay. A cyberattack on hospitals using robotic surgical devices could cause them to malfunction while in use, resulting in fatal injuries. If a country or terrorist group decided to take out a sitting U.S. senator undergoing robotically assisted surgery and then covered its tracks, the perpetrator’s identity would be hard to pinpoint, and there would be no clear U.S. legal precedent for classifying the hacking of hospital equipment as an assassination or an act of war. Nor do there appear to be clear protocols for retaliation.
There are less direct potential vectors of attack, too. Recently, a cold storage facility for embryos in Cleveland failed to notice that a remotely accessible alarm on its holding tanks had been turned off, leading to the loss of more than 4,000 frozen eggs and embryos. Many operators of industrial control systems never bother to change all their default passwords or security credentials, which can leave them vulnerable to ransomware attacks, and even fewer health care officials are likely to assume that someone might deliberately shut off sensors that monitor the viability of future human life. It is difficult to determine whether the Cleveland eggs and embryos were lost due to a simple maintenance failure or deliberate tampering—but as techniques such as the freezing of eggs become more common in wealthy nations, such a simple attack could wipe out thousands of future citizens...
...Cyberattacks—some egregious, some mundane—are happening now, quietly and unnoticed by the public. Much of the confusion and fear over cybersecurity comes from the distorted publicity surrounding a few outlying events. While cybersecurity experts can’t have perfect certainty over attribution or even the existence of some attacks, we can understand the larger security landscape, in which cybersecurity is merely a banal and predictable component of national infrastructure. The risk of cyberattacks is knowable, probabilistically.
Technology and cyberspace are changing faster than countries can legislate internally and negotiate externally. Part of the problem with defining and evaluating acts of cyberwarfare against the United States is that U.S. law is unclear and unsettled when it comes to defining what constitutes an illegal cyberact as opposed to normal computer activity by information security researchers.
The legal status of most information security research in the United States therefore remains unclear, as it is governed by the poorly drafted and arbitrarily enforced 1986 Computer Fraud and Abuse Act (CFAA)—a piece of legislation that was roundly derided by tech experts on its inception and has only grown more unpopular since. The law creates unnecessary fear that simple and useful information security research methods could be maliciously prosecuted.
These methods include network scanning using tools such as Nmap (a computer network discovery and mapping tool) or Shodan (a search engine for devices on the internet of things) to find unsecured points of access to systems. Such scanning does not constitute the exploitation of computer or network vulnerabilities; a real-world equivalent would be walking down a street and noting broken windows, open doors, and missing fence planks without actually trespassing on someone else’s property. One of the fastest fixes for the dismal state of federal cybersecurity expertise would be to overturn the CFAA and reward cybersecurity researchers engaged in preventive research instead of tying their hands with fears of breaking the law. Yet at present the U.S. government ham-handedly discourages many information security researchers from entering what should be a noble service.
This dynamic has left the U.S. government with critical shortfalls in top-level information security experts. The United States simply lacks a viable legislative plan for hardening its infrastructure against cyberattacks and developing much-needed cybertalent. Any strong defense against cyberattacks should follow the same principles used for basic U.S. infrastructure design: strategists plan, technicians execute, and experts examine. For example, the interstate highway system in the United States, authorized in 1956 to enable rapid military transport of troops and supplies, also had much broader civilian benefits...
...Cybersecurity should be akin to a routine vaccine, a line item in the infrastructure budget like highway maintenance. Basic cybersecurity measures—such as upgrades to encryption, testing the capability of recovery in the event of data loss, and routine audits for appropriate user access—should be built into every organizational budget. When incidents happen—and they will happen as surely as bridges collapse—they should be examined by competent auditors and incident responders with regulatory authority, just as major incidents involving airlines are handled by the National Transportation Safety Board (NTSB).
At present, however, the United States lacks an NTSB for cybersecurity. Due to the government’s lack of expertise, it is overly reliant on large companies such as EY, PwC, and Deloitte to handle this work. If the U.S. government isn’t capable of running a post-mortem on major cyberevents, citizens should be asking why—instead of letting lawmakers hand the work to contractors. Responding to major cyberattacks requires battalions of highly trained government analysts, not armies of accountants and attorneys...
Tarah Wheeler is an information security researcher and political scientist. She is a
New America cybersecurity policy fellow and senior director of data trust and threat and vulnerability management at Splunk. @tarah
No comments:
Post a Comment