Real talk: The imminent and very real danger of IoT

The last few years have seen a surge in everyday devices coming online: coffeemakers, refrigerators and beer kegs, to name a very few, each with their very own IP addresses. And those are just within your kitchen. Collectively called the Internet of Things (IoT), these devices are indeed game-changers.

But it’s no wonder Gartner informs us that in its Hype Cycle for Emerging Technologies, 2016, IoT is currently at the peak of inflated expectations. Manufacturers are more than happy to fulfill consumer demands to connect everything to the Internet.  

Mirai strikes, once, twice, thrice

As it always happens, however, in the rush to win customers, ease of use or time to market trumps safety and security. This was abundantly clear on October 21, when US and European internet traffic was significantly disrupted by a Distributed Denial of Service (DDoS) attack on Domain Name System provider Dyn. This attack impacted at least 80 major websites, including Twitter, Spotify, Reddit, Github and Netflix. Mirai malware, which has at its disposal a botnet army consisting of 500,000 IoT devices, was responsible. It used only 10% of the devices available in its network to mount this massive attack.

(Mirai was behind another DDoS attack in September 2016, after which its developer released the malware's source code to the hacker community, giving the larger community an opportunity to create their own network. The October 21 attack was Mirai's third.)  

POTUS weighs in

The disruption was big enough that everyone took notice. President Barack Obama asked, "How do we continue to get all the benefits of being in cyberspace but protect our finances, protect our privacy?" US senators, Homeland Security and other security officials called for more robust cybersecurity for IoT devices. Experts called it a wake-up call, noticing the ease of hacking the IoT devices due to lack of security features.

At least one Chinese electronics device maker, Hangzhou Xiongmai, recalled some of its components that go in surveillance cameras for a security fix.

Real talk on IoT security

Due to lack of security features, creating an IoT botnet is a great deal easier than phishing users to compromise PCs. Given the ease with which IoT devices can be hacked, we can expect more attacks to follow. Mirai, Japanese for ‘future,’ has given us a view into the future through these attacks, which include data breaches and ransomware attacks through compromised IoT devices.

Many of these attacks could be targeted at consumers of IoT. Imagine your viewing history on your DVR being stolen, IoT light usage monitored, thermostats manipulated, and coffee machine or refrigerator usage monitored. Is your privacy and safety impacted? Would this be on dark web for nefarious purposes? I joke that the market for classic air-cooled 1970s Porsches or Volkswagens will skyrocket as a response to concerns around data collected from connected cars and car hacking.

To some it may seem like science fiction, but Mirai has merely scratched the threat surface and given us a glimpse of things to come.

 FICO fights IoT, and other, cyber threats

My view is that the ever-improving artificial intelligence that FICO has been using to fight fraud for the last two decades, as well as new machine learning algorithms for self-learning, are part of the solution. For enterprises, deploying machine learning and artificial intelligence-based cybersecurity solutions is critical to protect enterprise IoT devices, making them self-aware of manipulation.

Further, continuous monitoring of connected systems by self-learning AI at network and end-point levels is a must for real-time detection of compromises. Monitoring and determining usual behavior of the devices, and taking countermeasures when behavior starts looking abnormal, is important to stop compromise, stop the transmission of private data, or stop its use in attacks on infrastructure. 

The ability to learn which behavioral changes are legitimate, and which aren’t, already plays a pivotal role in FICO's cybersecurity offerings. FICO’s proven self-learning analytics, artificial intelligence, machine learning-based algorithms and anomaly detection techniques are used to monitor activity across networks and real-time data streams. These technologies identify threats as they occur while maintaining low false positive alarm rates even for new types of threats.   

It’s time to get real about IoT security. Now is the time for our devices to get smarter. Read more about my IoT security concept, the “intelligence of things,” here. And follow me on Twitter @ScottZoldi