Many organizations facing a data-security incident struggle to understand how or why their organization was targeted in an attack. Most simply believe they are too small or too obscure to be targeted by malicious cyber actors. Even larger, well-known businesses are lulled into complacency, mistaking years without a major security incident as evidence that their business is an unlikely target, or believing that a small corner of their business, perhaps the new cloud instance they’re testing, will go unnoticed. They reason that with bigger or more prominent fish in the ocean, their relative obscurity is a strong line of defense. But this reasoning misunderstands how victims of cyber attacks typically become victims, and how easy it is for attackers to find and compromise vulnerable targets across the internet. While some victims are targeted for a specific purpose, especially by nation-state actors, many are not. More often they are opportunistic victims or victims of collateral damage directed at others. Understanding how attackers target victims is critical to proper network defense and to accurately assessing an organization’s risk scenarios.
Targets of opportunity
This article highlights four common ways attackers find their victims: phishing campaigns, watering hole attacks, malvertising, and mass internet scans for vulnerable or open systems. Knowing how these techniques find victims is key to understanding that every organization with internet-accessible systems and data is a target.
Now one of the most widely known attack methods, phishing campaigns blanket the internet with emails containing malicious attachments or links. They target individuals and organizations of all sizes, and often worm their way from one victim to the next using victims’ compromised accounts and address book contacts to spread, making the phishing campaign that much more effective because malicious messages arrive from a trusted or known source. The purpose behind many of these attacks is to harvest users’ credentials for sale or use in follow-on attacks. Other phishing campaigns with malicious attachments install ransomware; malware that collects banking details, credentials or other sensitive data; or remote access tools, signaling to the attacker when a catch is on the line.
Despite increased awareness and training, this attack method remains popular because it works, and our experience demonstrates that any organization can fall victim to these attacks.
Watering hole attacks
A second way attackers find targets of opportunity is through a so-called watering hole attack, where attackers compromise a legitimate website (the watering hole), inject malicious code into the site and then wait for victims to visit it. When a victim arrives at the site, the injected code checks the visitor’s browser and plug-ins for vulnerabilities and exploits them in the background. These attacks sometimes target specific industries or categories of individuals (targeting government contractors, for example, by compromising a site covering a related subject), but they also often target the general population by compromising widely visited pages like news sites. Victims are often surprised to learn these attacks can compromise a user’s system as soon as the victim visits the compromised site. The user doesn’t have to click on a link, download anything or take other action – when a user visits the compromised site with a vulnerable browser or plug-in, the attack is complete. Adobe Flash is a well-known plug-in with a history of serious and frequent vulnerabilities. Despite these known vulnerabilities and plans to retire the plug-in next year, we see frequent examples of attackers targeting this plug-in to compromise users’ systems.
Malvertising (malicious advertising) is a close cousin of the watering hole attack, but instead of compromising the site that unsuspecting victims visit, the attacker buys advertising space on the legitimate site and populates the purchased space with an advertisement containing malicious code. When the malicious advertisement loads into a user’s vulnerable browser, the code compromises the user’s system behind the scenes, again with no additional action by or notice to the user.
Mass internet scans
Perhaps the most opportunistic way attackers find victims on the internet is through mass scans of the entire internet for publicly exposed vulnerabilities. Although internet-wide scans may sound impractical or sophisticated, several tools and services automate the process, making it easy for even unsophisticated attackers. A tool released last month named AutoSploit, for example, combines the power of the Shodan search engine (which scans the internet and catalogs exposed systems and open ports) with the Metasploit penetration-testing framework (which contains a database of known exploits and automates the exploitation process) to search the internet for vulnerable systems and automatically exploit them. Other tools automate the discovery of servers with open remote-desktop protocol (RDP) services, the discovery of cloud storage locations like Amazon S3 buckets that customers have failed to secure or the presence of sensitive access keys mistakenly published in public repositories like GitHub.
This means that if your organization places a server or an IoT device on the internet with an insecure service or known vulnerability, or places sensitive data in an improperly configured S3 bucket or other cloud storage location, it’s only a matter of time before an attacker (or security researcher) discovers it.
Attackers who find these vulnerabilities leverage them in many ways. Some infect compromised systems with ransomware or develop other ways to extort victims; some steal sensitive data to sell or otherwise monetize through fraud; others sell the RDP or cloud-service access they compromise to buyers who use that access for other criminal activity (including cryptocurrency mining and using the compromised servers as hop points, or proxies, to conceal the true source of their malicious activity). An avalanche of news stories and my firm’s experience dealing with scores of clients victimized in these ways confirm these are real scenarios that organizations of all sizes face every day.
Collateral damage from attacks targeting others
Aside from the opportunistic attacks described above, several examples from the past year demonstrate how targeted attacks by sophisticated actors can affect a wide range of unintended and unexpected victims. Two prominent examples are the May 2017 WannaCry outbreak and the June 2017 NotPetya attacks. WannaCry, which authorities have recently attributed to the North Korean government, masqueraded as a common ransomware attack, but was, in fact, a purely destructive attack meant to disrupt its victims. Although it’s unclear whether the North Korean government unleashed WannaCry on a particular target, we know the virus spread across the globe rapidly, infecting targets of opportunity along the way and causing significant damage.
Similarly, the U.S. and U.K. governments recently attributed the NotPetya attack to the Russian government and its desire to destabilize Ukraine’s financial, energy and government sectors. Despite this limited motive, the virus spread to businesses around the world, causing significant collateral damage to many other organizations, including the law firm DLA Piper and the Danish conglomerate A.P. Moller-Maersk.
We also saw in 2017 a supply-chain attack affecting the popular privacy tool CCleaner. Here, the suspected nation-state attacker’s apparent motive was espionage against 12 technology companies. Yet, despite this limited motive, the attack compromised CCleaner’s legitimate update process to inject malicious code into approximately 2.27 million end-users’ systems. Although the attacker delivered the attack’s second-stage malware into only a small fraction of limited targets, every user that downloaded the malicious update was infected with the first-stage malware and exposed to further compromise.
These examples demonstrate how sophisticated attacks by nation-state actors can affect organizations of all sizes, even if the affected organization is not an intended target.
Assessing the risk
The scenarios described above have the potential to create serious legal, regulatory and reputational risk. Each scenario that impacts the confidentiality, integrity or availability of personal data, however briefly, is a “breach of security” and an internally recordable incident under the European Union’s General Data Protection Regulation (GDPR) Article 33(5). Depending on the data affected and the risk of harm to individuals, the incidents may also trigger notification obligations under GDPR Articles 33 and 34, and under various U.S. state and federal breach-notification laws. (The GDPR takes effect in May and includes extraterritorial provisions that apply to organizations operating outside the EU in certain circumstances.)
Moreover, an organization that fails to implement reasonable and appropriate security measures to address these risk scenarios could expose itself to increased litigation and regulatory risk. GDPR Article 32, for example, requires data controllers and processors to “ensure the ongoing confidentiality, integrity, availability and resilience of processing systems and services” after assessing “the risks that are presented … to personal data transmitted, stored or otherwise processed.” Likewise, various state and federal regulators may view failing to address these risk scenarios as unfair or deceptive business practices, or violations of regulations mandating reasonable security measures (e.g., the HIPAA Security Rule or the N.Y. Department of Financial Services Cybersecurity Requirements).
And because many scenarios outlined above result in mandatory notification obligations or public disclosures by security researchers, these incidents create reputational risk for organizations that may lose customers’ – or, perhaps more important, business partners’ – confidence in their operations. Especially under the GDPR’s new requirement that processors notify controllers of personal data breaches “without undue delay” (see Article 33(2)), repeated notifications of even small incidents could cause significant reputational damage and business loss.
Therefore, understanding how these attacker techniques translate into opportunistic attacks and collateral damage to all connected organizations is a critical first step to properly evaluating an organization’s legal, regulatory and reputational risk scenarios. And the variety and complexity of these attacks once again reinforces why organizations undertaking risk assessments must work with internal or external assessors with a strong understanding of these threats.