171 CDS 11 E rev. 1 final - INFORMATION AND NATIONAL SECURITY
LORD JOPLING (UNITED KINGDOM), GENERAL RAPPORTEUR
II. THE INFORMATION AGE AND THE NOTION OF SECRECY IN INTERNATIONAL RELATIONS
III. DIGITAL (H)ACTIVISM
IV. CYBER ATTACKS AND CYBER DEFENCE
V. INFORMATION AND CYBER SECURITY: OPTIONS FOR THE INTERNATIONAL COMMUNITY AND NATO
1. The ongoing information revolution poses a series of political, cultural, economic as well as national security challenges. Changing communications, computing and information storage patterns are challenging notions such as privacy, identity, national borders and societal structures. The profound changes inherent in this revolution are also changing the way we look at security, often in unanticipated ways, and demanding innovative responses. It is said that because of this revolution, the time it takes to cross the Atlantic has shrunk to 30 milliseconds, compared with 30 minutes for Intercontinental Ballistic Missiles (ICBMs) and several months going by boat.1 Meanwhile, a whole new family of actors are emerging on the international stage, such as virtual “hactivist” groups. These could potentially lead to a new class of international conflicts between these groups and nation states, or even to conflicts between exclusively virtual entities.
2. One of the most fundamental characteristics of the Information Age is its ability to connect. In this regard, the main tool is the Internet and the fact that bandwidth and storage capacity is currently doubling every 12 months.2 Interconnectivity is now central to government offices, critical infrastructures, telecommunications, finance, transportation, and emergency services. Interconnectivity is also central to culture and education. Even where communication and data exchanges are not routed through the Internet, they still, in many cases, use the same fibre optic cables.3
3. Despite its inherent advantages, this dependence on information technology has also made state and society much more vulnerable to attacks such as computer intrusions, scrambling software programmes, undetected insiders within computer firewalls, or cyber terrorists. The Internet is inherently insecure as it was designed as a benign enterprise of information exchange, a decentralised patchwork of systems that ensures relative anonymity. It is ill-equipped to trace perpetrators or to prevent them from abusing the intrinsic openness of the cyber domain. In this context, the key national security dilemma of the Information Age is how to create an effective and transparent government, which, at the same time, is also able to protect its citizens and vital national interests. Furthermore, in this Information Age, the North Atlantic Alliance faces a dilemma of how to maintain cohesion in the environment where sharing information with Allies increases information security risks, but where withholding it undermines the relevance and capabilities of the Alliance.
4. It is a critical time for the NATO Parliamentary Assembly (NATO PA) to discuss cyber security, as the Alliance has recently adopted its new comprehensive Cyber Security Policy and Action Plan. The details of this document are not publicly available for understandable reasons. Since the cyber domain is extremely dynamic and increasingly complex, cyber security and defence strategies of the Alliance as well as of individual Allies will be in a constant need of updating and revisiting.
5. This report will focus on three facets of the linkage between Information Age and national security. First, it will discuss the changing notion of secrecy in international relations. This issue was brought to prominence by the so-called “Cablegate” scandal. While the publication of classified diplomatic correspondence was not a result of a cyber attack, it is nevertheless directly linked to the information revolution: remarkable advances in data storage technology allowed one person to easily download colossal volumes of data that has taken the print media months, and possibly years, to digest and to publish.
6. Second, the explosion of Internet usage is creating the phenomenon we refer to as “digital (h)activism”. Social media and other Internet-based communities are creating new, ad hoc and cross-border allegiances that can manifest themselves in a variety of positive (reinforcing civil societies in authoritarian countries) and negative (empowering hacker groups that act against those who do not share their political worldview) ways.
7. Third, the report will discuss the challenge of direct cyber threats against states and, in particular, NATO’s role in cyber defence as one of the principal topics for the Euro-Atlantic community, particularly in the wake of the Lisbon Summit.
8. The report will not address the specific issue of cyber crime. While cyber theft and child pornography are issues of grave concern for the international community4, they do not have direct national security implications and are addressed by a number of other international organisations, including the UN, EU, OSCE, OECD and G8. The Council of Europe Convention on Cybercrime – which requires its parties to criminalise a number of activities in cyber space relating to infringements of copyright, computer-related fraud and child pornography – is a particularly noteworthy initiative that has yet to be ratified by several NATO member states.5
9. This report also represents the continuing effort by the Committee on the Civil Dimension of Security to discuss the issue of critical infrastructure protection within the Alliance. Cyber technologies are not only key enablers for systems such as energy generation or transport, but can themselves be considered as critical national infrastructure.
10. The report also builds upon the contribution by other NATO PA Committees, particularly the 2009 Sub-Committee on Future Security and Defence Capabilities report NATO and Cyber Defence [173 DSCFC 09 E bis] by Sverre Myrli (Norway) and the 2007 Science and Technology Committee report Transforming the Future of Warfare: Network-Enabled Capabilities and Unmanned Systems [175 STC 07 E bis] by Sen. Pierre Claude Nolin (Canada).
11. This chapter will discuss the challenges of protecting classified information in the age of Internet. It will also outline the political and security implications of the “Cablegate” scandal that highlighted the inter-agency and international co-operation versus sensitive information security dilemma.
A. THE “CABLEGATE”
12. According to the 11 September attacks investigation, the US government failed to ensure adequate information sharing, which could have prevented the attacks (FBI failed to share details connected to an al-Qaeda operative, who later proved to be key in uncovering the plot). As a result, representatives of the political elite, the military, and the financial world all pressed for wider sharing of classified information in order to increase operational efficiency in protection of the country. Therefore, the US government adopted a policy of information sharing, which it applied to numerous US governmental institutions and agencies including the Department of Defense (DoD) and the State Department (DoS).
13. This policy resulted in an exponential number of people obtaining access to classified information. Approximately 854,000 people now possess top-secret security clearances.6 For almost 10 years now, embassy cables have been distributed through the SIPRNet (Secret Internet Protocol Router Network operated by the DoD), which has made them accessible to DoS employees all around the world, to all members of the US military and contractors with necessary security clearance. Eventually, several millions of people ended up having access to materials such as US diplomatic cables.7 According to information-security experts familiar with the SIPRNet, the data-sharing system was not programmed to detect unauthorised downloading by anyone who had access to this pool of data. Thus, those in charge of the network design relied on those who had access to this sensitive data to protect it from abuse. These users were never scrutinised by any state agency responsible for the data-sharing system.8
14. The US government’s post-9/11 policy on information-sharing received the most serious blow when the “anti-secrecy” organisation WikiLeaks started publishing documents of different levels of confidentiality. Its first major release (April 2010) was a video of a US helicopter shooting into a crowd in Bagdad in 2007 which killed 18 people, including two Reuters journalists. Shortly after, the release of 77,000 documents allegedly revealing the realities of the Afghan war were made public, as well as almost 400,000 secret Pentagon documents on the Iraq war.9 In November 2010, WikiLeaks started releasing about 250,000 US diplomatic cables, many of which were classified. The cables provided US diplomats’ candid assessments of terrorist threats and the behaviour of world leaders.10 Currently, the US authorities suspect that the material was leaked by Private Bradley Manning stationed in the Persian Gulf, who had downloaded the information from a computer in Kuwait. He then allegedly passed these files on to the “whistleblower” organisation, which made them public.
B. REACTION TO THE LEAKS
15. WikiLeaks has spurred public debate with each of its releases. Nevertheless, the November 2010 release of US diplomatic cables got the most aggressive reactions from politicians world-wide. In anticipation of the leaks, Secretary of State Hillary Clinton and her diplomats warned foreign officials about the upcoming leak days before the November 2010 release happened. Following the release, the White House11 as well as the DoS were quick to denounce the leak and, as Secretary of State Clinton put it, characterised the cable disclosure as an “attack on both the United States and the entire international community”.12 Consequently, countries including Turkey, Iraq, Afghanistan, China as well as NATO were quick to condemn the leak.13
16. On the day of the release, the White House ordered government agencies to review security procedures and ensure that only the necessary users had access to their documents.14 Soon after, the President’s Office also appointed an Interagency Policy Committee for WikiLeaks, which was to assess the damage caused by the leaks, co-ordinate agencies’ reactions, and improve the security of classified documents.15 The US DoD conducted an internal 60-day review of security procedures. It also disabled the usage of different storage media and the capability to write or burn removable media on DoD classified computers.16 The Defense Information Systems Agency has also launched a new Host-Based Security System, which is meant to monitor software and policy rules in order to spot suspicious behaviour and alert responsible authorities. For example, the software should set off an alarm if large quantities of data are being downloaded. Today, approximately 60% of SIPRNet is protected by the software. In order for it to be bullet-proof, however, it will probably require additional compartmentalisation of information.17 A similar tracking mechanism is being adopted by US intelligence agencies (referred to as “enhanced automated, on-line audit capability”).18
17. The DoS has limited the number of people with access to the Net Centric Diplomacy database, which contains diplomatic reports19 suspended the access to SIPRNet and to two classified sites ClassNet and SharePoint, as well as prohibited the use of any removable data storage devices.20 Following the leaks, the US Air Force has blocked its employees’ access to at least 20 websites containing the leaked documents such as “The New York Times” and “The Guardian”. The Pentagon prohibited its employees to access the WikiLeaks website on government computers “because the information there is still considered classified”.21 Eventually, the administration banned hundreds of thousands of federal employees of the Department of Education, Commerce Department, and other government agencies from accessing the site. The Library of Congress, one of the world’s biggest libraries, also issued a statement saying that it would block WikiLeaks.22
18. As far as the WikiLeaks website was concerned, following the leak it suffered repeated distributed denial of service attacks, which prompted it to move its server. Companies such as Visa, Mastercard or Paypal suspended all their services to the organisation, which relies heavily on online donations from its supporters worldwide.23
C. TRANSPARENCY VS. SECRECY
19. The relationship between transparency and secrecy remains a key dilemma in the Information Age and has dominated worldwide media, especially since the outbreak of the WikiLeaks phenomenon. On the one hand, there are pro-transparency advocates who argue that the existence of WikiLeaks certifies that transparency of governments and other organisations is publicly desired. According to them, it is precisely the current Internet age that is conducive to institutional reform, increases public trust in government conduct, and enhances co-operation.24 And, as transparency proponents argue, we should not react to this development by limiting the spread of technologies and information, but instead by focusing on adapting the conduct of diplomacy, military affairs and intelligence to the new paradigm.25
20. That said, the Rapporteur believes that even if one is in favour of transparency, military and intelligence operations simply cannot be planned and consulted with the public. Without some secrecy, it would be impossible for governments, and especially security agencies, to perform their functions and to protect citizens. Added to which, transparency can be misused on several levels – by providing unprofessional or poor-quality interpretation of information or documents, by lack of experience on the topic or by pursuing a political agenda. Thus, not everything carried out under the “transparency label” is necessarily good for the government and its people. Moreover, the very ideal of transparency can also force public figures to become more secretive. The Information Age and its transparent nature may, for example, prevent diplomats from conducting “business as usual” such as making off-the-record statements or engaging in frank discussions with their colleagues.26 It also increases pressure on decision makers, who have to identify, assess, and react to information, which is immediately and widely accessible to other governments, organisations, as well as the public.27 This is an unnecessary and possibly dangerous pressure, especially when it comes to the issues of security.
21. This chapter will discuss the phenomenon of emerging borderless communities and networks, most of which are welcome, but some of which are highly dangerous. Virtual communities operating on-line provide new opportunities for civil society, but they have also increased the potential for asymmetrical attacks.
A. THE PHENOMENON OF HACTIVISM
22. Apart from causing harm, destruction or conducting espionage, some of the most recent cyber attacks have also been used as a means to reach a rather different goal. “Hactivism” is a relatively recent form of social protest or expression of ideology by using hacking techniques. Hactivists use different malware (or “malicious software”) and Distributed Denial of Service (DDoS) attacks to publicise their cause rather than for crime. Such attacks first occurred in 1989 but have gained more prominence over the last decade. In the past hactivists have attacked NASA, the Indonesian and Israeli governments, Republican websites, as well as the University of East Anglia.28
23. One of the most prominent group of on-line hackers - Anonymous - led a campaign against Iran, Australia and the Church of Scientology.29 Their most prominent campaign, however, took off in 2010 after WikiLeaks had released the US diplomatic cables. In its on-line seven-point manifesto, Anonymous announced its engagement in “the first infowar ever fought” and named PayPal as its enemy.30 What followed were DDoS attacks against Mastercard, Visa, PayPal, and other companies that had decided to stop providing services for WikiLeaks (they used to administer online donations for the site), against the Swiss bank PostFinance, that had earlier closed Julian Assange’s bank account, and against the Swedish Prosecution Service.31 The group also attacked Amazon.com, which was previously renting server space to WikiLeaks.32
B. THE ROLE OF THE SOCIAL MEDIA
25. The discourse on the Information Age and new social media gained a new momentum in the beginning of 2011, as numerous countries in North Africa and the Middle East began experiencing popular anti-government uprisings. It was the Internet, in combination with other new and old media such as cell phones and television that has enabled global resistance to authoritarian rule in the region. The sight of protesters holding up signs “Thank you, Facebook!” has become common in Egypt and Tunisia.34 Journalists, experts and politicians are increasingly using terms such as “Facebook Revolution”, “Twitter Diplomacy”, or “Cyber-Activism”.35 Today, Facebook is a community that unites more people than any other country in the world, save for China and India, and if the growth trends keep going as they are, the social network site will soon have more users than India has inhabitants.36
26. Social media, and most prominently Facebook, have helped activists in many of these countries to organise anti-government protests, evade surveillance, discuss issues that have been taboo for decades such as torture, police violence or media censorship, and provided a platform for trading practical tips on how to stand up to rubber bullets and organise barricades.37 Recognising that new social media have had an important share in the success of public resistance, two days after demonstrations started in Egypt, Facebook, telephones, and Internet all over the country were switched off. A few days later, when the Internet connection was restored and Facebook users regained access to their accounts, they found that the regime attempted to use this tool for disseminating pro-Mubarak propaganda. Most recently, Facebook pages, groups and blogs attempting to mobilise protesters have appeared in Algeria, Bahrain, Morocco and Syria. As a show of support for the protestors, the online group Anonymous attacked websites of the Tunisian and Egyptian government, Mubarak’s National Democratic Party and the Tunisian stock exchange, making them unavailable for certain periods of time.38
27. Proponents of social media argue, that “merely knowing about social dynamics changes social dynamics”. The authority of one’s peers has been proven to have substantial influence on the decisions made and thanks to these new social media peer influence has evolved into multiple, nation-wide protests. However, others argue that the influence of new social media in respect to the 2011 revolutions has been overrated. Critics say that social media can only provide fast coordination of masses but do not deliver the narrative or resolve that are essential for starting and sustaining any popular movement.39 As an example, in Egypt the protests started growing significantly after the government had shut down the Internet. The social media also do not prevent popular protests from being contained by governments and security services. In other words, they do not determine the outcome.
28. In the wake of popular uprisings in North Africa and the Middle East, social media representatives have reacted very differently to the events. Facebook’s representatives declined to discuss Facebook’s role in the uprisings and provided only a short statement: “We’ve witnessed brave people of all ages coming together to effect a profound change in their country. Certainly, technology was a vital tool in their efforts but we believe their bravery and determination mattered most.” Twitter and YouTube (owned by Google), embraced their roles in the protests more openly. As opposed to Facebook, they took a proactive approach after the Internet was shut down in Egypt by assisting protesters in setting up a new service, "speak2tweet", that would allow people to communicate and organise.40 WikiLeaks founder, Julian Assange, was even more eager to attribute the success of these recent resistance movements to his site. According to him, it was the US diplomatic cables leaked by WikiLeaks that gave the army ‘the confidence that they needed to attack the ruling political elite’..41
29. Most recently, in June, Europe’s last dictatorship was also struck by a wave of antigovernment rallies. Due to severe shortage of dollar and euro reserves, the Belarusian government devaluated its national currency, which resulted in overnight pressure on living standards. As a consequence, opponents of these measures started anonymously organising themselves through social networking sites such as Facebook and its Russian equivalent vKontakte.42 After initial arrests, organisers opted for so-called “silent” forms of protest. By posting instructions on-line, they called on people to fill up parks or squares without doing anything apart from clapping their hands, having their phones buzz or play music at an agreed time, or simply drive slowly through Belarusian towns playing the popular Soviet-era song called “We Are Waiting for Change”.43 So far the state police have been unable to identify those posting instructions via social media. The new concept of “silent demonstrations” is making it difficult for the police to know who is actually taking part in the protest. The demonstrations have not yet managed to mobilise large numbers of supporters or pose any real threat to the ruling elites. They have, however, managed to utilise social media to involve several thousand people of all professional backgrounds as well as different age-groups.44
30. As mentioned above, the Information Age has brought about an environment that has made the state and society more vulnerable to digital attacks. They are vulnerable because we no longer keep our files and data on a shelf, but in a virtual world accessible from any of the world’s corners. As in the case of WikiLeaks, these files can be physically removed from a computer, handed over to adversaries, or simply made public. Apart from that, however, one of the greatest strengths as well as weaknesses of the Information Age is that files can also be accessed and on-line services disrupted from afar by various “cyber attacks”. The term “cyber attack” represents a myriad of activities ranging from stealing passwords, to accessing accounts, disrupting critical infrastructure of a country or spying on an enemy.45 As cyber experts testified to the members of two NATO PA Sub-Committees during the visit to The Hague on 18-20 April 2011, there is still no agreement within the international community as to which of these cyber activities constitute a crime. NATO C3 Agency’s Principal Scientist Brian Christiansen suggested that the existing legislative “black holes” should be addressed in a multinational manner due to the transnational nature of the threat and this argument has been supported by many cyber security specialists.46
31. Due to its decentralised nature, the Internet per se is in fact extremely robust and resilient as it was designed to withstand nuclear war. However, separate parts of this network of networks are vulnerable to cyber threats. The most disquieting feature of the cyber domain is that the attacker has the advantage over the defender. Perpetrators need only one weak point to get inside the network, while defenders have to secure all vulnerabilities. These attacks also take place at the speed of light which leaves little or no time react to attacks. Furthermore, the inherent nature of the Internet allows an attacker to forge the sender’s address or to use botnets (zombie computers often located in different countries), thereby disguising the true identity of an attacker and leading to misattribution of the source of an attack.47It is estimated that roughly 1,200 botnets reside on US soil alone.48
32. The problem of attribution is widely recognised as the biggest obstacle for effective cyber defence. Professional hackers can easily cover their tracks and thus avoid penalties. Deterrence, a critical element of a traditional defence paradigm, is problematical in cyber space. In addition to which, most cyber attacks are performed by civilian hacker groups so it is very difficult to prove government involvement. For instance, experts suggest that the thriving Chinese hacker community is not directly supervised by respective government authorities but merely encouraged financially or through ‘patriotic’ education mechanisms such as the People's Liberation Army's militia and reserve system. It makes it difficult to blame Beijing for the attacks such as the one in 2007, when some 25-27 terabytes of information (equivalent to roughly 5,000 DVDs) were illegally copied from the Pentagon.49
33. According to Kenneth Geers of the NATO Co-operative Cyber Defence-Centre of Excellence (CCDCOE) in Estonia, who spoke on the issue at the CDS Committee’s Spring Session in Varna, Bulgaria, 27-30 May 2011, the internet and computer programmes are so complex that they are almost impossible to secure. It is not, however, entirely impossible to track down cyber attackers. Firstly, in order to attribute successfully cyber attacks, we need to develop a system of international cooperation among governments and experts, possess a network of effective human intelligence and conduct thorough police investigations. All these steps are essential because simply outlawing hacking or only employing cyber means when tracing attacks is insufficient. According to Kenneth Geers, NATO, being a powerful alliance of members with high tech capabilities and great financial assets, is the right organisation to tackle the issue. Secondly, we might be able to solve the problem of attribution thanks to the new Internet Protocol version 6 (IPv6), which has built in authentication technology and makes it possible to limit interaction only to confirmed ID whitelisted – users into your network. This technology limits internet connectivity but, on the other hand, it provides a new level of protection.
34. At the moment, however, sources of cyber attacks are almost impossible to trace. Nevertheless, when it comes to the involvement of states in cyber attacks, Russia and China are said to be the usual suspects.50 From what we know today, terrorist groups such as al-Qaeda do not yet have the capability to carry out such attacks. However, terrorists are increasingly acquainted with the possibilities offered by the Internet. Extremists have long used the Internet to spread their ideals as well as details of tactics, techniques and procedures used in terror attacks. Since 2001, many internet sites have been monitored and shut down in the United States.51 But sites are constantly changing and security officials need to be agile in locating them. Furthermore, chat-rooms and online publications are used not only to spread their violent message amongst supporters but to radicalise and recruit new members also. Of note is al-Qaeda’s ‘Inspire’ web publication which was reportedly recently hacked itself by British security officials.52
35. As noted, the cyber domain is extremely dynamic and rapidly developing, making it difficult for cyber security experts to always react adequately and speedily to novelties. For instance, one of the newest trends is the emergence of the so-called ‘cloud computing’. Cloud computing is network-based computing where software, data storage and other resources are provided over a shared network. It allows users to access their company’s business applications securely through the ‘cloud’.53 Governments as well as the private sector benefit from cloud computing, which helps to increase productivity, cut costs (according to Brookings Institute’s estimates, the US government can save up to 25-50% of its IT costs), keep pace with technology innovation, and become more transparent with their citizens.54 Nevertheless, the process also raises some key data security concerns, which include: vendors using ineffective security practices, agencies not able to examine the security controls of vendors, cybercriminals targeting data-rich ‘clouds’, and agencies losing access to their data if the relationship with a vendor ends.55 Thus, standards to regulate this new cyber space need to be set and implemented.56 According to Gregory Wilshusen, Director of Information Security issues at the US Government Accountability Office (GAO), however, US agencies are moving their data to the ‘cloud’ before government-wide security strategy has been developed by responsible agencies. As he continued, “these risks generally relate to dependence on the security assurances and practices of a service provider and the sharing of computing resources.”
36. There are, however, also voices that believe cloud computing will improve security. According to Mike Bradshaw, Director of Google Federal, “Cloud computing vendors store data on multiple servers in multiple locations, making it difficult for cybercriminals to target one location”. Also, vulnerabilities can be managed more rapidly and uniformly.57
A. TYPES OF CYBER ATTACKS
37. Generally speaking, there are two types of cyber attacks: Distributed Denial of Service (DDoS) and malware attacks.
1. DDoS attacks
38. DDoS attacks aim to overwhelm a target by sending large quantities of network traffic to one machine. Attackers take over a number of other computers (botnets) and use them without the knowledge of their owners – for instance, the Estonia attack, roughly one million computers were hijacked in 75 countries.58 The goal of DDoS is to prevent legitimate users from accessing information and services, such as the actual computer, email, websites, online accounts (banking, etc.). DDoS attacks are extremely difficult to deal with because they do not attempt to exploit vulnerabilities of a system. Vulnerabilities may be patched, but essentially one cannot do much to prevent DDoS attacks.59
39. One of the first major attacks aimed to cripple a country’s critical infrastructure hit Estonia in May 2007. The e-government country experienced co-ordinated DDoS attacks on websites of the Estonian President and Parliament, almost all of its government ministries, political parties, major news organizations, two banks and several communication companies. The attacks came soon after Estonian authorities had relocated a Soviet war memorial in Tallinn – a step which spurred protests by ethnic Russians living in Estonia. The series of cyber attacks, which occurred weeks after the event, supposedly originated in Russia and were hosted by Russian state computer servers. Russia denied these allegations, but in March 2009, an activist with the pro-Kremlin youth group Nashi claimed responsibility for organising the cyber attacks on Estonia. It should be noted that Estonia is extremely dependent on the Internet. At the last parliamentary elections, ¼ of the voters cast their votes via Internet.
40. Another significant DDoS attack was launched against Georgia in the summer of 2008. This is of note due to the fact that it was coupled with the use of conventional military force, something that a number of experts predict will occur more often in the future. Georgia blamed Russia for the attack only for Russia to deny any involvement.60 A year later, the combination of cyber and conventional force was supposedly also employed in the case of the bombing of the Syrian nuclear reactor, which was allegedly orchestrated by Israel.61
2. Malware attacks
41. Malware – or “malicious software” – attacks refer to techniques capable of infiltrating one’s computer without the user’s knowledge and taking control of it, collecting information, or deleting its files (see examples of malware in the Annex). Attack malware can reportedly be bought online for several hundred dollars or even downloaded for free.62
42. Malware-based cyber attacks are increasingly being used for espionage. In 2008, the Unites States experienced a major attack on the classified networks of US Central Command in charge of oversee military operations in the Middle East and Central Asia. Based on available information, the attack was carried out by a foreign intelligence service, which used portable data storage devices to spread malware. In 2009, the GhostNet cyber espionage study conducted by the Information Warfare Monitor concluded that 1,295 computers in 103 countries, had been penetrated by GhostNet malware that allowed the surveillance and possible control of states’ critical cyber infrastructures. Worryingly, 30% of GhostNet’s targets were classified as high value.63
43. Espionage cyber attacks, however, can also be carried out against non-state actors such as private companies and think tanks. “Operation Aurora” carried out in late 2009/early 2010 is a case in point. Over several months, Chinese hackers managed to penetrate the networks of at least 34 financial, technological, and defence companies by exploiting flaws in e-mail attachments.64 One of the attack’s targets, the giant search engine Google, admitted that hackers had penetrated Gmail accounts of Chinese human rights advocates in the United States, Europe and China. A number of human rights organisations and Washington-based think tanks focusing on United States-China relations were also hit by the attacks. According to experts, the attack reached a new level of sophistication as hackers exploited multiple flaws of different software programmes – multiple types of malware codes were allegedly used against multiple targets and the whole process was very precisely co-ordinated. This series of attacks was aimed at gaining information about the latest defence weapons systems, source codes powering software applications of prominent technological companies, as well as gaining background about Chinese dissidents.65
44. The Stuxnet is technically a malware, but its characteristics, originality and potential for disruption are so novel that it merits special attention. The Stuxnet worm has been described as “the most sophisticated cyber weapon ever deployed”66 and its widely-acknowledged role in damaging Iran’s Bushehr nuclear reactor and Natanz uranium enrichment plant has put Stuxnet firmly in the spotlight recently.67 Essentially, the worm is a direct-targeting cyber attack: it “sniffs” around its target’s operating system and only attacks if this system matches its targeting criteria, thereby making detection harder for other defences. Once it has acquired its target, Stuxnet deploys two extremely complicated programming payloads to “bomb” them. In the Iranian example, the first of these cyber bombs attacked the centrifuges in the nuclear plant, slowly "unsynching" them so that they collided with each other, causing serious damage. The second cyber bomb compromised the digital warning, display and shut-down systems controlling the centrifuges, thereby blinding these systems to the reality of what was happening.
45. This characteristic makes Stuxnet unique in that it specifically attacks and compromises the Supervisory Control and Data Acquisition (SCADA) systems of critical national infrastructures. Thus, the real danger of Stuxnet is that, although the Iranian example was a specifically targeted attack, the same method could be used to virtually attack any information technology system used in any critical infrastructure around the world. Stuxnet has therefore been described as a “cyber weapon of mass destruction”.68 Of particular note is that the vast majority of complicated information technology systems controlling critical national Infrastructures that are potentially vulnerable to Stuxnet are located in NATO and NATO partner countries. Related to this, British Telecom has estimated that 65 % of cyber attacks on critical infrastructures exploit pre-existing configuration errors in the controlling system’s software, highlighting the need for standardisation across the Alliance.69
B. NATO AND CYBER DEFENCE
1. NATO’s cyber agenda
46. The cyber domain is often described as the “fifth battlespace”; representing both opportunity and risk for the military. In the context of the revolution in information and communication technology, the military institutions of major powers have been working relentlessly to interconnect commanders, soldiers, sensors and platforms in order to improve agility and achieve better situational awareness. Today, more than 1/5 of US defence and security acquisitions are in the cyber sector.70 “Network-centric capabilities” has become a buzzword in militaries, while new technologies enable commanders to make better-informed decisions and to reduce human losses by, for example, operating an unmanned aerial vehicle (UAV) over Afghanistan from a base in Nevada.
47. On the other hand, our armed forces are now faced with risks they have not experienced before, such as the incident reported by The Wall Street Journal in December 2009, when Iraqi insurgents managed to intercept feeds coming from American UAVs using inexpensive software that is available on the Internet.71 The Pentagon computer systems are probed up to six million times per day, according to US Cyber Command.
48. NATO’s increasing involvement in cyber security is therefore inevitable. As NATO Secretary General Anders Fogh Rasmussen put it: “[t]here simply can be no true security without cyber security”. The Alliance has included this issue on its agenda since 2002 when it approved a Cyber Defence Programme – “a comprehensive plan to improve the Alliance’s capability to defend against cyber attacks by improving NATO’s capabilities”. However, it was not until the 2007 attacks against Estonia that NATO embarked upon developing a comprehensive cyber defence policy that would include not only the protection of the Alliance’s own networks but would also augment the cyber security of individual member states. The Group of Experts’ Report (the "Albright report") recommended that NATO must accelerate its efforts to respond to the dangers of cyber attacks. It recommended focusing on protecting NATO’s communications and command systems, helping Allies to improve their ability to prevent and recover from attacks, and developing an array of cyber defence capabilities aimed at effective detection and deterrence.
49. At the Lisbon Summit, NATO member states committed the Organisation to developing a revised NATO Policy on Cyber Defence that was adopted by NATO Defence Ministers in June 2011, together with the Action Plan that sets out the details of implementing the Policy. The contents of the Policy remain classified, but, according to the official NATO press release, the Policy addresses all key aspects relating to the Alliance’s cyber security, including bringing all NATO structures under centralised protection, clarifying NATO’s response mechanisms to cyber attacks, integrating cyber defence into NATO’s Defence Planning Process, devising the framework of assisting national efforts of individual Allies, facilitating better information sharing and setting up principles of closer co-operation with non-NATO countries, international organisations and the private sector. This Policy will most likely require regular revisions and updating as the developments in the cyber domain are remarkably frequent.
50. At present, individual members continue to bear the principal responsibility for the security of their networks, while relevant NATO structures, apart from protecting their own networks and providing support for NATO operations, are expected to assist member states by sharing best practices and dispatching Rapid Reinforcement Teams in case of emergency. At present NATO cyber efforts are purely defensive in nature, and there is a particular focus on protecting member states Critical National Infrastructures.
51. Key NATO institutions in the area of cyber security include:
52. NATO conducts annual exercises aimed at enhancing the understanding of NATO’s cyber defence capabilities and identifying areas for improvement. This year’s exercise, Cyber Endeavor was scheduled to take place from 5-22 September 2011 in Grafenwöhr, Germany.
53. A lot remains to be done, however. NATO’s principal cyber unit – NCIRC – is only partially operational and does not yet provide 24/7 security for all NATO networks. Full operational capability is expected to be achieved in 2012. NCIRC is also only engaged in passive defence, monitoring network activities and dealing with incidents.
54. With the NATO Policy on Cyber Defence being classified, discussion continues on how NATO should react to cyber attacks against one of its member states. In particular, questions arise as to the relevancy and practicality of invoking Article 5 of the Washington Treaty in response to a cyber attack. The Washington Treaty refers specifically to “armed attacks”, but the New Strategic Concept is vaguer and the word “armed” is dropped in reference to collective defence. While this does not change the Washington Treaty, one can presume that the Alliance is more open to the idea of applying Article 5 if a cyber attack on member states were to cause significant casualties. However, questions still arise as to what response mechanisms the Alliance should employ against attackers. Should the retaliation be limited to cyber means only, or should conventional military strikes also be considered? Furthermore, the Alliance must decide to what extent it can engage in cooperation on sensitive cyber issues with partner countries, such as Russia.
2. National policies of member states
55. As noted above, member nations bear the principal share of responsibility for their cyber security. Before the 2007 attacks against Estonia, most European nations were developing national strategies to promote Information Society focusing on economic and cultural benefits offered by new communication and computing technologies, largely neglecting possible risks. Since 2007, the need for a more balanced approach has been increasingly acknowledged.72
56. The 2010 UK House of Lords report on cyber security noted wide differences between various European countries in terms of preparedness to meet cyber threats. Since in the cyber domain the system is as strong as the weakest link, the report stated that the European countries “have an interest in bringing the defences of the lowest up to those of the highest”.73 The exact level of preparedness is difficult to measure, however, due to a lack of full understanding of the complexity of the cyber domain.
57. The highest level of preparedness in the Alliance is in the United States and the United Kingdom. The United States feels more threatened by cyber attacks than any other nation due to its highly pervasive use of information and communication technology as well as to its status as a superpower. President Obama identified cyber security as a strategic priority. From 2010 to 2015, the US government is expected to spend over US$50 billion on its cyber defences.74 The Departments of Defense and Homeland Security share the responsibility for the security of American government networks and implement this mandate through several agencies such as National Security Agency and US Cyber Command (inaugurated in 2010 and specifically tasked to protect US military networks). In terms of legislation, three separate Acts streamlined executive responses to cyber warfare on critical national energy infrastructures, while another Act coordinated wider cyber security efforts, including those against financial institutions and industry.75 In July 2011, the Pentagon released its new Cyber Strategy (known as “Cyber 3.0”). The document considers cyberspace as an operational domain and focuses on “active defence”, i.e. strengthening traditional network protection measures with other capabilities such as signal intelligence. It is not clear, however, if the document empowers cyber defence institutions to go after an attacker. The new Strategy also emphasises closer interinstitutional, international as well as public-private co-operation.76 The Strategy, focusing on defensive measures, has also proved false the allegations that the United States was considering militarising cyberspace and prioritising development of offensive cyber weapons.
58. The UK’s leading cyber agency is the Government Communications Headquarters (GCHQ). Cyber security occupies a central place in the National Security Strategy and the Strategic Security and Defence Review published in October 2010. Experts note that the “review contains all the early signs of a well-balanced and (now) better-funded approach to UK cyber security.”77 The UK Computer Misuse Act is also hailed as “a robust and flexible piece of legislation in terms of dealing with cybercrime”.78
59. That said, even in the United States and UK there are still important questions that need to be addressed. In particular, experts note the insufficient degree of cooperation between the government agencies and private sector which owns most of information capabilities and infrastructure – more than 90% of American military and intelligence communications travel through privately-owned telecommunications networks.79 However, private entities are reluctant to allow greater government involvement and monitoring. The UK House of Lords report noted that representatives of the commercial United Kingdom Internet industry showed little interest in giving evidence for this report. Many experts stress that private industry makes its decisions on cyber security measures based on financial rather national security calculations.
60. While the United States and the UK tend to lead on these matters, other NATO members have also updated their existing legal frameworks and made cyber security increasingly prominent in their security strategies. In particular, significant progress has been achieved in establishing Computer Emergency Response Teams (CERTs). A CERT is an organisation that studies computer and network security in order to provide incident response services to victims of attacks, publish alerts concerning vulnerabilities and threats, and to offer other information to help improve computer and network security. The 2010 House of Lords report identified the lack of CERTs in some European countries as a major concern. However, in 2011 the situation seems much better. According to the register of the European Network and Information Security Agency (ENISA), CERTs were established in all European NATO countries. Furthermore, the establishment of more advanced Computer Security and Incident Response Teams (CSIRTs) is being promoted. CSIRTs are CERTs that have extended their services from being a mere reaction force to a more complete security service provider, including preventive services like alerting and security management services.80
61. However, there is no basis for complacency. Establishment of new institutions must be followed by more intensive schedule of joint exercises. The legislative basis must also be further reviewed and updated to take into account the new realities of the cyber domain. According to NATO Deputy Assistant Secretary General Jamie Shea, legislative frameworks in many NATO countries are lagging behind in cyber term realities.81 At the meeting with NATO Parliamentarians in The Hague on 19 April 2011, NATO C3 Agency General Manager Georges D’hollander said that not all NATO member states have adopted legislation that would make it mandatory for the private sector to protect their data and their networks. For instance, it should be mandatory to install safeguards that would prevent computers or networks being hijacked and used as ‘botnets’. NATO C3 Agency’s Principal Scientist Brian Christiansen also suggested that all NATO nations should employ the so-called “red teams” that use hackers’ methods to probe security levels of various national networks (without malign intentions, of course).
62. The less advanced NATO nations must realise that in the cyber domain there cannot be a free ride. One study notes that nations that do not have adequate legislative and institutional framework to protect their cyber assets are less likely to receive assistance from the international community because “in a rapid reaction situation, existing procedures better support effective interaction (…) because there is a certain amount of ‘homework’ that can only be performed by the victim.”82
63. The challenges of the Information Age for national and international security are complex and require the combined efforts of international, regional and national authorities and the private sector, as well as sub- and trans-national groupings of active individuals. NATO is not in a position to address all aspects of this challenge, but it does have a significant role to play, not least because it unites nations with the most developed information and communication infrastructure (infrastructure, hardware and software which collectively make up the Internet are still overwhelmingly Western designed and produced; more than 50% of the world's Internet traffic transits the United States).83
64. On the global level, NATO should support initiatives to negotiate at least some norms of acceptable behaviour for the cyber domain. This framework must discourage the cyber arms race and clearly prohibit the use of cyber attacks against civilian infrastructures. The principles of international law should also recognise indirect responsibility of a state to ensure that its territory is not used by non-state actors to launch attacks against a third country. If a country systematically fails to ensure that or provides sanctuary for perpetrators, it should be considered as breaching international law and should face sanctions.84 When addressing our Committee at the Assembly session in Varna, Kenneth Geers of the NATO CCD COE suggested that the universal cyber treaty could follow the path of the Chemical Weapons Convention, i.e. focus on promoting best practices, helping find data points quickly, and sending teams to collect forensics, and eventually securing networks.
65. Achieving this agreement will not be easy, since some critical players – such as China – view cyber security from an “information security” perspective. This perspective is based on their desire to limit dissent and access to information deemed threatening to their regimes. These nations have proposed in-built tracking devices on all Internet packets that would allow all actions on the Internet to be traced. Western analysts argue this would be cumbersome, costly and easily negated by criminal groups, intelligence agencies and militaries. Therefore, the real target of such proposals is the average Internet user and their ability to access information and engage in political dialogue anonymously.85 Such a surveillance approach is prohibited by many NATO member states’ own laws governing surveillance, propaganda and counter-terrorism.
66. Other approaches to policing the cyber domain focus on developing technical solutions within Internet infrastructure itself to help maintain security. The Internet was originally designed to be interoperable and has therefore paid little attention to security aspects. The 2003 US National Strategy to Secure Cyberspace identified vulnerabilities within three “key Internet protocols”: the Internet Protocol, which guides data from source to destination across the Internet; the Domain Name System, which translates Internet Protocol numbers into recognisable Web addresses; and the Border Gateway Protocol, which provides the connection between networks to create the “network of networks”86. None of these protocols have in-built mechanisms to verify the origin or authenticity of information sent to them, leaving them vulnerable to being manipulated by malicious actors. Therefore, funding and developing technical solutions for a new set of secure protocols that will address many of the vulnerabilities in the current Internet infrastructure whilst falling short of surveillance of member states populations could be useful to NATO.
67. In addition, NATO member states should support wide ratification of binding international treaties, like the Council of Europe’s Convention on Cybercrime, because banning cyber criminal activities would also help negate cyber terrorists as well as state-sponsored cyber attacks that often use the same techniques as cyber criminals. The verifiability of these conventions is a serious issue, however.
68. In terms of public-private co-operation, relevant authorities of NATO nations should be more pro-actively engaging private IT companies when it comes to setting stricter rules on the use of cyber space. Dialogue is essential because software companies like Microsoft and Google remain able, by developing various software options, to exercise influence beyond what any nation state could aspire to do using their legislative powers. Incentives must be put in place to encourage private companies, particularly those running critical national infrastructures and designing cyber hardware and software, to upgrade their security systems beyond simple profit vs. loss calculations. It is also important for our nations to co-operate closely with Internet Service Providers in order to identify and quarantine the compromised computers (botnets) residing on their soil.
69. The Alliance should also establish closer co-operation with the EU based on already existing agreements. Although NATO is developing cyber defence capabilities, it still needs the EU because it issues laws on comprehensive standards for cyberspace and NATO does not. It would be useful, however, if the EU established the position of an EU “Cyber Czar” in order to have a clear contact point for NATO.
70. With respect to its own contribution, the most immediate objective for the Alliance is to ensure swift and efficient implementation of the newly adopted Cyber Security Policy and Action Plan. NATO should incorporate its cyber policies (and encourage its member states to do likewise) into a broader framework for adapting the military to the realities of the Information Age. Cyber security is not a value per se, it must be seen within the context of the developing concept of network-enabled capabilities. In other words, we need to find the right balance between the advantages offered to our armed forces by the new information and communication technologies, and the introduction of stricter protective measures against cyber threats, measures that could result in reduced efficiency of the military.
71. It also goes without saying that NATO must clarify its response mechanisms for itself in case of a cyber attack against one or more of its members, although these mechanisms do not necessarily need to be announced publicly in order not to let the adversaries know what they could get away with. Some argue that Article 5 should not be applied with respect to cyber attacks because their effect so far has been limited to creating inconvenience rather than causing the loss of human lives and because it is hard to determine the attacker. So far, there is no evidence that cyber attacks took human lives. However, the Rapporteur believes that the application of Article 5 should not be ruled out, given that new developments in cyber weapons such as Stuxnet might eventually cause damage comparable to that of a conventional military attack.
72. In more practical terms, NATO should consider its role in protecting physical infrastructure associated with the cyber domain. The physical vulnerability of fibre-optic cables and information hubs represent a serious challenge within the cyber domain. Most long-haul fibre-optic cables reach land at obvious choke points, which make them susceptible to attack or damage. Of note is the choke point for transatlantic cables, Widemouth Bay, Cornwall, in the UK, where four major EUUS cables reach land.87 This area has reportedly been designated “vital to US security” because of these cables.88 Meanwhile, the vast majority of the physical cables that connect the United States and Asia run through the Luzon Strait choke point between Taiwan and the Philippines.89 Cables in the Malacca Strait are also congested, and island NATO members and partners, like Iceland, Japan and Australia, are particularly vulnerable.90 To date, the best form of protection for , these sub-surface cables has been their anonymity. However, sometimes this is not enough, as highlighted by the fact that 75% of Internet capacity between Europe and a large part of Asia was temporarily lost when, in 2008, ships off the Egyptian coast severed two inter-continental fibre-optic cables by dragging their anchors.91 A Georgian woman denied 90% of Armenians access to the Internet for 5 hours when she inadvertently cut through a cable with her spade.92 There have also been other large Internet disruptions caused by cable incidents in Malta, Sicily, the United States and Asia.93 These highlight the possibility of sabotage by state or non-state actors. In terms of bandwidth capacity, NATO member states are heavily dependent on infrastructure in the United Kingdom for their transatlantic communications. Much of these key Internet peering points are based in and around London and have previously been threatened by flooding.94 Any disruption to these infrastructures could have far-reaching economic and military effects.
73. Other elements of NATO’s better preparedness against cyber attacks include further strengthening of national cyber incident response teams, achieving full operational capability of NCIRC, intensification of joint exercises, promoting more efficient sharing of best practices among the Allies and a wider use of “red teams”. Before investing in highly elaborate cyber defence systems, however, the Allies should first ensure that proper levels of basic “computer hygiene” are routinely maintained.
74. Security of networks in critical national infrastructure objects should remain a key priority. Technical solutions being examined in this regard include the introduction of high fidelity sensors to monitor intrusion activity on networks, and the strengthening of fault tolerance techniques.95 However, for a truly comprehensive cyber approach to infrastructure resilience, technological solutions alone will not suffice. A collaborative approach between citizens/systems users, businesses, law enforcement agencies and civil institutions will provide the best cyber security for these objects.96
75. The Rapporteur also suggests that NATO considers applying common funding procedures for procurement of some critical cyber defence capabilities for its member states. The Alliance and its nations should also redouble their efforts to invest in human capital, because currently the Western nations are widely believed to be losing their advantage in cyberspace in terms of numbers of cyber experts and qualified personnel.
76. Other practical measures should include reviewing our policies in terms of critical information that is to be stored online. The “Cablegate” revealed some documents that date back to 1966. Nigel Inkster, a prominent British expert, says that this “suggests an excess of zeal among those tasked to place State Department data on SIPRNet, since these cannot be relevant to today's operational requirements.” It is also necessary to review the operating systems of critical national infrastructure with a view to limiting their unnecessary exposure to online connections. Furthermore, new safeguard mechanisms must be put in place to prevent unauthorised downloading of sensitive data to digital storage devices. Procedures for vetting relevant personnel should also be revisited.
77. That said, the Rapporteur wishes to emphasise that all necessary security measures should not cross the line where they would violate the fundamental principles and values cherished by the nations of the Euro-Atlantic community. It is also important for our national security interests: since the cyber domain is to a large extent governed by the people, it is important to win the moral support of the majority of the virtual community. In order to prevent abuse by the governments, stricter security rules should be accompanied by measures ensuring democratic oversight. For instance, the United States announced recently the establishment of the Privacy and Civil Liberties Oversight Board (PCLOB) to ensure that privacy and civil liberties are protected.97
78. Last but not least, the Rapporteur would like to underline the role of parliamentarians not only in terms of issuing relevant legislation, but also in communicating with a public that is often insufficiently informed about the scope of opportunities and risks posed by the Information Age.
Types of Malware
Creates a “back door” into a computer, which can be obtained via the Internet from anywhere around the world. It can delete, steal or monitor data on someone else’s computer. It can also turn the computer into a “zombie” and use it to hide the real perpetrator’s identity and cause further damage to other systems. 98
Monitors and keeps track of keystrokes on a computer usually without the user being aware of it. The information can be saved to a file and sent to another computer. Acquiring private data such as usernames and passwords are usually the key targets of the programme.
Infects files when they are opened or being run and is capable of selfreplication. It often manifests itself as a logic bomb or a Trojan. Viruses are difficult to track and can spread very quickly. In 2000 the ILOVEYOU virus caused damage of approximately US$10 million.