The Internet of Things: A Trifecta of Cyber and Physical Threat Risks
The recent WannaCry ransomware cyberattack provided another chilling reminder of the potential disruptive power behind the Internet of Things. Even before the WannaCry attack in May 2017, a distributed-denial-of-service (DDoS) attack on a domain name server provider, Dyn, Inc., took place in October 2016, pushing many popular internet services offline for hours. The Dyn attack, which utilized the malware Mari as the supporting agent, was a sea-change event carried out by hundreds of thousands of internet-connected devices, such as routers, security cameras and DVRs, that rely on default factory user names and passwords coupled with weak or nonexistent security protections. It illustrated that hackers can now target vulnerable low-hanging fruit and turn it into a super botnet to carry out the DDoS attack. One takeaway from the Dyn attack is that the exponential growth of devices coming online, some 5.5 million per day according to Gartner, creates an unparalleled ecosystem for malevolent actors to find and weaponize the Internet of Things (IoT).
The WannaCry ransomeware attack shut down UK hospitals, Russian computers, factories, and multiple businesses and personal interests around the globe. It has not yet been attributed to a specific actor, although North Korea has been identified as a potential perpetrator. According to The Wall Street Journal, Kaspersky Lab ZAO said the malware appeared in 74 countries. Later reports placed the number of countries impacted at 150.
Japan was among the countries that felt the impact of the attack. The Nikkei Asian Review reported 2,000 terminals and 600 IP addresses had been hit. Among these was a computer for the water and sewer services in the city of Kawasaki. The article went on to note, “the use in infrastructure of connected devices, part of the Internet of Things, made room for the attack.”
Hiroki Takakura, a professor at the National Institute of Informatics, in the same article noted: “Production and control devices and other equipment are made to match the systems they are used with, so it can be difficult to update them. The attackers targeted systems that still run on outdated operating systems such as Microsoft Windows XP. More users, lately, are unable to apply the latest security updates due to such issues as software incompatibility, which is something of an alarm bell for Internet of Things.”
Assessments & Warnings
In the aftermath of the WannaCry attack, cybersecurity experts again emphasized the vulnerabilities of weak or insecure IoT devices. In a posting, 5 Security Predictions for a Post-WannaCry World, Nicole Henderson provided predictions shared by cybersecurity company eSentire. One of those identified was worm-based attacks that could do physical damage: “worm based attacks could unleash physical damage to infrastructure as we move to the Internet of Things.”
With the two latest demonstrations, stakeholders need to address security concerns from the ground up. The trifecta of threats has already been documented and demonstrated. The IoT is vulnerable to attacks that can (1) cause physical damage to persons and property, (2) cause a widespread distributed denial of service to servers and computer systems, and (3) deny access to computer systems to secure ransom.
On May 11, 2017, Director of National Intelligence Daniel R. Coates appeared before the Senate Select Committee on Intelligence to provide the U.S. intelligence community report on Worldwide Threat Assessment. A section of the report, “Emerging and Disruptive Technologies,” identified IoT as one such technology:
“The widespread incorporation of ‘smart’ devices into everyday objects is changing how people and machines interact with each other and the world around them, often improving efficiency, convenience, and quality of life. Their deployment has also introduced vulnerabilities into both the infrastructure that they support and on which they rely, as well as the processes they guide. Cyber actors have already used IoT devices for distributed denial-of-service (DDoS) attacks, and we assess they will continue. In the future, state and non-state actors will likely use IoT devices to support intelligence operations or domestic security or to access or attack targeted computer networks.” (Emphasis added.)
The United States Government Accountability Office (GAO) issued a Technology Assessment report for the Internet of Things in May 2017. Among the challenges the GAO identified are information security, privacy, safety, the need for standards and disruptive economic issues from the growth of the Internet of Things. The GAO noted it “…has previously reported that cyber threats to internet based systems are evolving and growing. Without proper safeguards, these systems are vulnerable to individuals and groups with malicious intentions who can intrude and use their access to obtain and manipulate sensitive information, commit fraud, disrupt operations or launch attacks against other computer systems and networks. The threat is substantial and increasing for many reasons, including the ease with which intruders can obtain and use hacking tools and technologies.”
The GAO identified the following types of attacks against the IoT:
- Denial of service
- Distributed denial of service
- Passive wiretapping
- Structured query language injection (SQLi controls a web application’s database server)
- Wardriving (search for Wi-Fi wireless networks by a person in a moving vehicle)
- Zero-day exploits (software tool that attacks a flaw in a computer system with no opportunity for detection).
The GAO noted that the lack of security controls in many IoT devices occurs in part because many vehicles, equipment and other IoT-enabled devices were built without anticipating threats associated with internet connectivity or the requisite security controls. Furthermore, the GAO noted that many IoT devices are configured with identical or near identical software and firmware that can magnify the impact of successfully exploiting technical vulnerabilities common to all of them.
While recommending that IoT devices should be designed with software update capabilities, the GAO recognized that IoT devices often are designed without software upgrade capabilities or with a cumbersome upgrade process. In addition, many IoT devices may be deployed with anticipated service lives many years longer than typically associated with high tech equipment, making it unlikely that security updates will continue throughout the entire service life.
The Department of Commerce Internet Policy Task Force & Digital Economy Leadership Team published a report in January 2017 − Fostering the Advancement of the Internet of Things − that noted:
- “…cybersecurity best practices are a new concept for many IoT stakeholders. Mature manufacturers of newly wired devices, such as an appliance manufacturer … may have little to no experience collecting, securing, and protecting consumer data…”
- “…start-ups building IoT technologies and interfaces for the first time may focus primarily on getting a product to market, without considering how to protect and secure computer networks or data.”
- “…different sets of best practices will be relevant for different IoT entities, such as hardware manufacturers/integrators, developers, deployers, and operators.”
The report also identified specific areas that may require special consideration, such as devices used by children and autonomous vehicles, noting that “just as there is no easy description for IoT itself, there is no single prescription for IoT security.”
So what are some takeaways from the most recent events?
First, the stakeholders perpetuating the growth of the IoT need to focus on security. The Department of Commerce (DoC) provides a good outline of the broad-based steps that need to be undertaken by all stakeholders looking to capitalize on the potential of the IoT. These include the need for flexible, risk-based solutions. In other words, threats and vulnerabilities are constantly evolving; therefore, predefined solutions become obsolete without the creation of cutting-edge solutions.
Second, there must be security by design, not as an afterthought. The approach needs to be holistic and take into account risk assessment during design and testing of products before they are deployed. Vulnerabilities discovered after the product leaves the manufacturer must be addressed with patching and support throughout the life cycle of the product. Regarding addressing technical limitations, the DoC report notes that many IoT devices have “…computationally weak hardware, minimal operating systems and/or limited memory…” Lightweight encryption may be one pathway to an answer for devices with limited computing power.
Just as there is no free lunch, there are no straightforward, surefire ways to address security vulnerabilities in internet-connected devices. Strategic Principles for Securing the Internet of Things from the U.S. Department of Homeland Security echoes many of the recommendations advanced by other federal agencies, including the Federal Trade Commission, the Government Accountability Office and the Department of Commerce. It starts with incorporating security at the design phase, promoting security updates and vulnerability management, building on recognized security practices, prioritizing security measures according to potential impact, promoting transparency across the IoT, and connecting carefully and deliberately.