Developing technologies such as 5G networking, the internet of things (IoT), and artificial intelligence (AI) are enabling everything from self-driving cars to natural language recognition. But these promising technologies are also creating new cybersecurity challenges for IT security professionals.
“There are several evolving attack vectors, but the three that really keep me up at night are 5G networks, the internet of things, and AI or machine learning,” said Paul Mazzucco, Chief Security Office for TierPoint.
Mazzucco explained how these emerging three attack vectors will impact cybersecurity in a recent webcast, Bots Evolve to Challenge Security in a 5G World.
Cybersecurity challenge #1: 5G & IoT enable more attack vectors
5G networks are rolling out in most major cities, and Verizon predicts that half of the U.S. will soon be using 5G. Besides faster speeds, 5G offers the ability to define use-case based virtual networks made to meet the range of needs of IoT devices. For example, a self-driving car may send and receive volumes of data for analysis, whereas an environmental sensor might transmit only temperature and humidity changes. 5G enables many customized networks over existing infrastructure. That capability is expanding the potential use cases for IoT.
There are already an estimated 27 billion IoT devices, in industries ranging from manufacturing, transportation, and telecommunications to healthcare, office buildings, and consumer homes. Forecasts suggest that that the number of IoT devices will reach 50 billion by 2030.
Unfortunately, many IoT devices have weak security, made weaker by careless users who often neglect to change the factory default passwords on their devices. The growing number of unsecured IoT devices means more opportunities for cyberattackers who can use them to infiltrate corporate networks and create botnets to launch denial of service (DoS) attacks.
Nor is 5G without its own security vulnerabilities. To create separate virtual networks, 5G uses software-defined networking (SDN). These multiple software-based network “slices” offer cyberattackers more targets to hack and increases the odds they can identify individual types of traffic.
Likewise, 5G uses short-distance transmissions, making it dependent on multiple, small cell towers located near sending organizations. That, again, expands the number of potential targets and helps hackers pinpoint which towers are most likely to transmit a particular organization’s traffic.
“A hacker can sit down next to the infrastructure, sniff out those packets, and attack particular ones,” said Mazzucco, citing hospitals with their volumes of personally identifiable information (PII) as high-value targets.
Challenge #2: Smarter bots can also exploit these attack vectors
AI is making everything smarter, so it’s no surprise that AI is also helping hackers create smarter crime bots. Bots are bits of code that do basic tasks. Unlike botnets which are networks of IoT devices infected with bot malware, individual bots are programs that do things like, search the web, scrape information off web sites, and retweet news items on social media. An emerging 4th generation of bots is leveraging AI to mimic human behavior — and doing it well enough that many security solutions fail to detect them.
“Hackers are using artificial intelligence and machine learning to be more effective,” noted Mazzucco, noting that estimated 20% of Internet traffic is caused by malicious bots.
For example, 4G bots can be daisy-chained to perform sophisticated, automated attacks which are hard to detect because they so perfectly execute human behavior—right down to moving the mouse pointer in a random pattern much like a human scanning a web page.
“4th generation bots know how long to press on icons to mimic humans and will even misspell words on purpose to look more authentic,” notes Mazzucco. “We call it behavioral hijacking using advanced machine learning.”
AI-powered bots can also spider environments to find out what security protocols are in place, what the best attack vector is, and how should the attack be structured for optimum success.
There is a positive side to this. The software-based nature of 5G should make it easier to monitor traffic for specific threats and fine-tune security policies for different types of traffic. In addition, AI and automation will also help to shore up 5G security. For example, automated analysis of traffic patterns using AI can learn what is normal vs abnormal behavior for specific devices, applications, and end-users, and spot potential attack patterns. Regression testing can look at patterns over long periods of time, such as 90 days past, and identify related anomalies spaced days or weeks apart.
AI is also helping cybersecurity developers create 4th generation “white hat” bots to seek out crime bots and stop them before they can do damage. For example, machine learning-based technologies, such as intent-based deep behavioral analysis (IDBA), are being developed to more accurately and quickly identify bad bot behaviors.
Also read: Forbes Tech Council – Can 5G Networks Stand Up To 4th-Gen Bots?
Face cybersecurity challenges head-on
As IT technologies continue to advance, there will always be cybercriminals looking to exploit vulnerabilities. Whether they succeed depends on how seriously IT security professionals enforce good cybersecurity policies and keep their security defenses up to date. Our IT security services let you customize your security solution to safeguard each layer within your environment with a multilayered approach. This ensures all of your systems are protected from cyberthieves. Contact us today to learn more.
Want to know more about the challenges of 5G, IoT, AI, and bots? Watch the full webcast and read Mazzucco’s articles Can 5G Networks Stand Up to 4th Gen Bots and The Next Generation of Bot Attacks.