2016 saw a huge number and variety of cyberattacks, ranging from a high-profile DDoS using hijacked Internet-facing security cameras to the alleged hacking of party officials during the US election. We also saw a rising tide of data breaches, from organizations big and small, and significant losses of people’s personal information. With the year almost over, we’re pondering how some of those trends might play out in 2017.
Current and emerging attack trends
Destructive DDoS IOT attacks will rise. In 2016, Mirai showed the massive destructive potential of DDoS attacks as a result of insecure consumer IoT (Internet of Things) devices. Mirai’s attacks exploited only a small number of devices and vulnerabilities and used basic password guessing techniques. However, cybercriminals will find it easy to extend their reach because there are so many IoT devices containing outdated code based on poorly-maintained operating systems and applications with well-known vulnerabilities. Expect IoT exploits, better password guessing and more compromised IoT devices being used for DDoS or perhaps to target other devices in your network.
Shift from exploitation to targeted social attacks. Cybercriminals are getting better at exploiting the ultimate vulnerability – humans. Ever more sophisticated and convincing targeted attacks seek to coax users into compromising themselves. For example, it’s common to see an email that addresses the recipient by name and claims they have an outstanding debt the sender has been authorized to collect. Shock, awe or borrowing authority by pretending to be law enforcement are common and effective tactics. The email directs them to a malicious link that users are panicked into clicking on, opening them up to attack. Such phishing attacks can no longer be recognized by obvious mistakes.
Financial infrastructure at greater risk of attack. The use of targeted phishing and “whaling” continues to grow. These attacks use detailed information about company executives to trick employees into paying fraudsters or compromising accounts. We also expect more attacks on critical financial infrastructure, such as the attack involving SWIFT-connected institutions which cost the Bangladesh Central Bank $81 million in February. SWIFT recently admitted that there have been other such attacks and it expects to see more, stating in a leaked letter to client banks: “The threat is very persistent, adaptive and sophisticated – and it is here to stay”.
Exploitation of the Internet’s inherently insecure infrastructure. All Internet users rely on ancient foundational protocols, and their ubiquity makes them nearly impossible to revamp or replace. These archaic protocols that have long been the backbone of the Internet and business networks are sometimes surprisingly flaky. For example, attacks against BGP (Border Gateway Protocol) could potentially disrupt, hijack, or disable much of the Internet. And the DDoS attack on Dyn in October (launched by a myriad of IoT devices), took down the DNS provider and, along with it, access to part of the internet. It was one of the largest assaults seen and those claiming responsibility said that it was just a dry run. Large-scale ISPs and enterprises can take some steps to respond, but these may well fail to prevent serious damage if individuals or states choose to exploit the Internet’s deepest security flaws.
Increased attack complexity. Attacks increasingly bring together multiple technical and social elements, and reflect careful, lengthy probing of the victim organization’s network. Attackers compromise multiple servers and workstations long before they start to steal data or act aggressively. Closely managed by experts, these attacks are strategic, not tactical, and can cause far more damage. This is a very different world to the pre-programmed and automated malware payloads we used to see – patient and evading detection.
More attacks using built-in admin languages and tools. We see more exploits based on PowerShell, Microsoft’s language for automating administrative tasks. As a scripting language, PowerShell evades countermeasures focused on executables. We also see more attacks using penetration testing and other administrative tools that may already exist on the network, need not be infiltrated, and may not be suspected. These powerful tools require equally strong controls.
Ransomware evolves. As more users recognize the risks of ransomware attack via email, criminals are exploring other vectors. Some are experimenting with malware that reinfects later, long after a ransom is paid, and some are starting to use built-in tools and no executable malware at all to avoid detection by endpoint protection code that focuses on executable files. Recent examples have offered to decrypt files after the victim shared the ransomware with two friends, and those friends paid to decrypt their files. Ransomware authors are also starting to use techniques other than encryption, for example deleting or corrupting file headers. And finally, with “old” ransomware still floating around the web, users may fall victim to attacks that can’t be “cured” because payment locations no longer work.
Emergence of personal IoT attacks. Users of home IoT devices may not notice or even care if their baby monitors are hijacked to attack someone else’s website. But once attackers “own” a device on a home network, they can compromise other devices, such as laptops containing important personal data. We expect to see more of this as well as more attacks that use cameras and microphones to spy on households. Cyber criminals always find a way to profit.
Growth of malvertising and corruption of online advertising ecosystems: Malvertising, which spreads malware through online ad networks and web pages, has been around for years. But in 2016, we saw much more of it. These attacks highlight larger problems throughout the advertising ecosystem, such as click fraud, which generates paying clicks that don’t correspond to real customer interest. Malvertising has actually generated click fraud, compromising users and stealing from advertisers at the same time.
The downside of encryption. As encryption becomes ubiquitous, it has become much harder for security products to inspect traffic, making it easier for criminals to sneak through undetected. Unsurprisingly, cybercriminals are using encryption in creative new ways. Security products will need to tightly integrate network and client capabilities, to rapidly recognize security events after code is decrypted on the endpoint.
Rising focus on exploits against virtualized and cloud systems. Attacks against physical hardware (e.g. Rowhammer) raise the possibility of dangerous new exploits against virtualized cloud systems. Attackers might abuse the host or other guests running on a shared host, attack privilege models, and conceivably access others’ data. And, as Docker and the entire container (or ‘serverless’) eco-system become more popular, attackers will increasingly seek to discover and exploit vulnerabilities in this relatively new trend in computing. We expect active attempts to operationalize such attacks.
Technical attacks against states and societies. Technology-based attacks have become increasingly political. Societies face growing risks from both disinformation (e.g., “fake news”) and voting system compromise. For instance, researchers have demonstrated attacks that might allow a local voter to fraudulently vote repeatedly without detection. Even if states never engage in attacks against their adversaries’ elections, the perception that these attacks are possible is itself a powerful weapon.
What can organisations do to protect against new threats?
Unfortunately, many organizations still don’t have their security basics right. We offer six measures organizations should put in place to help keep more complex threats at bay.
Moving from layered to integrated security. Many organizations now possess multiple solutions that were once best-in-breed but are now too costly and difficult to manage. Moving towards integrated solutions where all components communicate and work together will help to solve this. For example, if malware knocks an endpoint’s security software offline, network security should automatically quarantine that device, reducing the risk to your entire environment.
Deploying next-generation endpoint protection. As ransomware becomes ubiquitous and endpoints grow more diverse, organizations must refocus on endpoint protection. But signature-based solutions are no longer enough on their own, and can miss zero-day attacks. Choose solutions that recognize and prevent the techniques and behaviors used in nearly all exploits.
Prioritizing risk-based security. No organization possesses the resources to systematically protect everything, and 100% prevention is no longer realistic. Clarify the risks associated with each system, and focus your efforts accordingly. Risks change fast: look for tools that track them dynamically, and respond accordingly. But make sure those tools are easy and practical enough to use.
Automating the basics. You can’t afford to waste time running the same reports and performing the same security tasks you always have. Automate wherever it can be done simply and easily, so you can focus scarce resources on serious risks and high-value tasks.
Building staff and process to deter and mitigate social attacks. Since social attacks now predominate, educating users and involving them in prevention is now even more important. Focus education on the threats each group is likeliest to encounter. Make sure it’s up-to-date: outdated guidance on topics such as phishing can be counterproductive, offering a false sense of security.
Improving defender coordination. Cybercrime is organized crime: defense must be organized, too. That means choosing tools and processes that eliminate barriers within your organization, so everyone can respond quickly to the same attack. It may also mean looking for legal and practical opportunities to collaborate with other companies and the government, so you can mitigate widespread attacks and learn from others’ postmortems.