Thefts of trade secrets through cyber espionage and cyber harassment through distributed denial of service attacks have been business hazards for years. Cyber security insurance policies to guard against these threats have been available since the 1970s. More recently, however, true cyber attacks have been capable of causing real-world harm. In FBI Director James Comey’s first Congressional testimony, he warned that cyber attacks surpass terrorism as the primary threat against the United States. This threat is by no means limited to the United States.
In 2010, malware researchers discovered a computer worm called “Stuxnet.” Further investigation revealed that the program would spread indiscriminately, but had a very specific target: the industrial control systems used in Iran’s nuclear program. The worm, it was reported, damaged industrial equipment and interfered with production processes in Iran, while, at the same time, it reported false data to the control systems so that they reported no errors to the operators. This discovery has been described as “an Oppenheimer moment in the history of hacking.”
Malicious programs like Stuxnet can interfere with the supervisory control and data acquisition (SCADA) software on industrial control systems, while giving their users the false impression that all is well. Stuxnet’s progeny could target the SCADA interfaces on air traffic control systems, on dams, on the electrical grid, or on the energy distribution or storage facilities, among others. Most dangerous would be attacks on the control systems of chemical, food processing, nuclear or water treatment plants.
This article addresses the legal consequences for insurers and insureds in the event of such an unprecedented, but possible, event.
At a Department of Homeland Security National Protection and Programs Directorate Cyber Security Insurance Workshop in November 2012, a federal official stated that “companies should not assume that the federal government will take responsibility” as the insurer of last resort in a catastrophic cyber attack, unless the government first “requires them to adopt a particular security solution prior to a successful attack.” Even then, “it’s unclear that the government would own the liability.”
In 2002, the Terrorist Risk Insurance Program Act (TRIA) became law. TRIA sought to provide “for a transparent system of shared public and private compensation for insured losses resulting from acts of terrorism,” to protect consumers, ensure the availability and affordability of terrorism risk insurance, stabilize markets, and “build capacity to absorb any future losses.” TRIA defines terrorism as a violent act dangerous to human life, property or infrastructure, occurring within the US (or to a US air carrier, vessel or diplomatic facility), committed on behalf of a foreign person or interest, to coerce the American people or United States Government. Crucially, the act must be certified as terrorism by the Secretary of the Treasury, a determination not subject to judicial review. Such a certification creates a private cause of action for property damage, injury or death. TRIA, however, expires on December 31, 2014. Neither TRIA nor the draft TRIA reauthorization bills before Congress, discuss cyber security, much less cyber attack. While some in the insurance industry believe the Department of Treasury has informally indicated that it would certify a cyber attack as terrorism covered by TRIA, this is not clear, and efforts are underway to make explicit such a guarantee.
Industry and insurance representatives disagree over whether existing cyber insurance policies would cover physical damage from Stuxnet-type attacks, whether stand-alone policies on SCADA systems are needed, or whether traditional casualty and property insurance would cover losses incurred in a cyber attack. Some general liability policies exclude cyber incidents from coverage.
A response by the United States government to the threat of cyber attack is beginning to take shape, which will impact businesses in as-yet unpredictable ways. In February, President Obama’s Executive Order 13636 warned that the “cyber threat to critical infrastructure continues to grow and represents one of the most serious national security challenges we must confront.” That same month, pursuant to Presidential Policy Directive 21, DHS identified infrastructure “where a cyber security incident could reasonably result in catastrophic regional or national effects…” DHS named 16 such sectors in a list that encompasses significant segments of the economy: the agricultural, chemical, commercial, communications, dams, defense, emergency services, energy, financial, government, healthcare, information technology, manufacturing, nuclear, transportation and wastewater sectors.
In addition, a “voluntary” cyber security framework will be issued by the Department of Commerce next February. After a meeting between President Obama and corporate chief executives this October, the White House stated that the framework being created by the National Institute of Standards and Technology (NIST) is “intended to raise the level of cyber security across the U.S. critical infrastructure,” and “will lay out a set of core practices for organizations to manage their cyber security risk.” The same day, the NIST released for comment a framework for businesses to use in identifying, protecting against, detecting, responding to and recovering from cyber threats. The goal is to establish standards and best practices. The framework highlights legal and regulatory requirements, as well as methodologies, for protecting privacy and civil liberties. A possible implication of these forthcoming standards and best practices is that industries which do not follow these guidelines will have not met minimum standards of care. Business will need to adapt to these guidelines, and insurers will need to become familiar with them.
When a cyber attack on a SCADA system occurs, then the parties to insurance litigation in the aftermath will have to grapple with a number of novel issues:
• A computer virus is difficult to attribute positively to an individual, and possible to misattribute, frustrating efforts to identify for the purposes of TRIA the foreign actor responsible. In addition, a Stuxnet-type attack can be launched from anywhere in the world, potentially complicating the requirement that the act take place in the US.
• Senior Iranian security officials have publicly singled out the US power grid as vulnerable to retaliation, in response for what has been described in the press as an American and Israeli effort to use Stuxnet to attack Tehran’s suspected nuclear arms program. Iran often acts through non-state actors, such as Hezbollah. Query whether, given the international political implications, a Treasury Secretary would be willing to certify, for the benefit of a civil action brought by a private party, that Iran committed an act of war against the United States through a proxy.
• The recent Executive Order suggests intelligence indicating the likelihood of a cyber attack on identified US public and private infrastructure. While the US government is involving business more closely in cyber security, intelligence reporting on this subject is sensitive by its nature, and shared by the government with certain industry partners in a highly sanitized fashion, when it is shared at all. It is not clear whether cyber security intelligence is shared with the insurance industry.
The United States and other governments expect cyber attacks on SCADA systems. Companies with potential exposure, including insurers, should consider both reasonable precautions as well as the potential loss coverage well before such an event occurs.