In 2017, an unnamed casino found that its data servers had been compromised and called on the aid of a security firm to help them find the culprit. Shortly afterwards, the surprising results of this investigation were reported far and wide: like the plot of an ill- conceived James Bond story, hackers had entered the casino’s network through an Internet-connected thermostat in a decorative aquarium.
Today’s organizations have a lot more to worry about than the old fish tank trick: this year, Gartner predicts that the number of devices connected to the Internet will reach 20.4 billion, setting a world record that will continue to climb for years to come. In our time, connected refrigerators, printers, TVs, and smart meters will provide points-of-entry for hackers with increasing frequency.
In the past, we’ve written about the security problems plaguing the current generation of IoT devices: just two years ago, researchers at the Black Hat and DEFCON security conferences showed just how bad the problem is by hacking dozens of devices in unique and novel ways. This begs the question: how did we get here? Why is IoT so difficult to secure, and what can organizations do about it?
Why IoT is A Supply-Side Problem
To explain the IoT security problem, we have told ourselves a plausible story sometimes repeated on this website: IoT is an inherent security risk, because increasing the number of Internet-connected devices in an organization also expands the attack surface available to malicious actors.
But – while there is truth to this story – it does not explain the sheer number of easily prevented security issues in business grade IoT.
According to Ponemon Institute, 51% of organizations acquire IoT products through a third party; meanwhile, 48% of organizations have been subject to at least one IoT attack, and that number is rising. As we will see, these two facts are not unrelated.
How Vendors Cheat on Security
In the lack of industry regulations incentivizing high security standards for IoT products, the incentive for vendors to make a quick profit by cutting corners drives sloppy development, lack of vulnerability testing, and quality control issues galore. The IoT market is in its “wild west” phase, as the PC market was three decades ago, and organizations must be wary who they work with.
The following tactics are some of the most common ways we find IoT vendors punting the responsibility for secure design from themselves to their customers.
- Quick Turnaround
By now, we have been talking about the “Internet of Things” for years, but the hype cycle isn’t over yet: because it is still cited as one of the best ways for organizations to modernize and take advantage of “big data,” the demand for IoT motivates companies to join the market as fast as they can with an often-questionable supply.
Vendors with no history in the IoT market may introduce products too quickly without an adequate development cycles, patch “IoT” features into their existing product lineup, or simply label existing devices as “IoT”. Practices like these lead to devices that not only suffer from general quality issues, but easily succumb to probing and attack.
- No Vulnerability Testing
Vendors are not immune to the lack of security awareness which impacts their customers. While it may be in their best, long-term interest to offer products with a high bar for security, it’s all-too-easy for vendors to skip a comprehensive vulnerability testing phase, opting instead to run down a checklist of features, if even that. Many companies lack the capabilities to test their products for security issues in the first place, and without regulations forcing them to do so, they simply won’t bother.
- Convenience at the Cost of Risk
When it comes to ease-of-access, what benefits IoT customers also benefits hackers. For the sake of convenience, vendors make design choices that exacerbate the vulnerability of their products: web interfaces, for instance, are the biggest target of IoT attacks – even those behind a firewall with NAT can be compromised. Likewise, the omission of two-factor authentication (2FA) and forced credential updates is a decision driven by form over function, when both features could thwart a huge number of IoT attacks.
- Corner-Cutting
Vendors frequently cut corners to make their products work as intended, and these tactics incur a high security risk. Because most IoT devices are embedded, they lack the power to perform data encryption or key negotiation. While these functions could be implemented with a dedicated security chip, most vendors won’t bother due to the added cost of production.
Similarly, when IoT devices lack adequate data storage – or any storage at all – vendors will connect them with the cloud and advertise this as a feature rather than a security liability. Rather than build dedicated customer support channels, vendors will add easily exploited backdoors into the device’s firmware. The list goes on and on.
- Poor Firmware
Speaking of backdoors in IoT firmware, the design of firmware is a major contributing factor to IoT security issues: few vendors will dedicate the time it takes to work out all the kinks before release; debugging systems used in the staging system of a device are often left in, allowing hackers to dump a huge amount of useful information.
Lack of testing may leave firmware vulnerable to buffer overflow, and the use of open-source platforms leaves an unprotected attack surface completely visible to attackers. The best vendors update their firmware on a regular basis to patch for newly discovered vulnerabilities, but this is a rarity.
- API Flaws and External Threats
From the outside, IoT integration with third-party apps through an application programming interface (API) seems like a great idea: but API flaws left by vendors open the doorway to attacks from malicious code hidden within seemingly innocuous applications. Researchers have also proven the possibility of DNS-rebinding attacks on IoT through a website, infected link, advertisement or malicious redirect. In the future, organizations may have to worry that their network will be infected every time their employees browse the Internet.
How to Avoid Bad Vendors
The IoT security gap remains one of the greatest threats to security across federal agencies. In response, legislators have discussed the idea of enforcing IoT regulations for some time, and NIST has produced IR 8259, a draft of recommendations for IoT manufacturers.
But until that happens, bad IoT vendors will persist, and organizations must practice due diligence to protect themselves. Here’s how to do that:
- Take inventory of the IoT products throughout your organization, alongside any devices connected to the Internet (organizations should be keeping inventory of all their IT assets as part of a comprehensive security strategy).
- Conduct a vulnerability assessment to discover the devices that constitute a real threat to your organization, and remediate the issue. This will also give you an idea which vendors to avoid moving forward.
- Be careful who you do business with: vet your vendors during the product acquisition phase (industry reputation, quality control, customer testimonials and quality of business). Show an equal amount of caution when expanding the capabilities of IoT devices through third-party software vendors.
Adopting a Threat-Based Mentality
While they have never been more serious than they are today, the risks of IoT have been understood for over a decade. If organizations have ignored them, it’s because they have adopted a checklist mentality: but following regulations to the tee won’t protect against threats that legislation doesn’t address.
In order to protect their data, revenue and customers, today’s organizations must take a proactive approach to security. With the help of vulnerability and penetration tests, cyber hunt and asset management, “cybersecurity” can mean a lot more than compliance: it can mean safety against malware and targeted attacks.
Take stock of your IT assets and fix vulnerabilities before NIST tells you to: with a DoD background, our world-class experts in governance, vulnerability testing and ethical hacking can help through technical consulting and federal security services. Contact us today!