IoT devices – estimated to soon number in the billions – are now cybercriminals’ top attack target, surpassing web and application services, and email servers, according to an app and cloud security product company.
The fifth volume of The Hunt for IoT report by F5 Networks notes that 13 Thingbots – which can be co-opted by hackers to become part of a botnet of networked things – were discovered in the first half of 2018. Six were discovered in 2017 and nine in 2016. Spain was the top country under attack during the past 18 months, including enduring a remarkable 80pc of all monitored IoT attack traffic between January 1 and June 30, 2018. Other countries under consistent pressure included Russia, Hungary, the US and Singapore.
Most of the attacks between January and June originated in Brazil (18pc of instances). China was the second biggest culprit (15pc), followed by Japan (9pc), Poland (7pc), the US (7pc) and Iran (6pc). The most infected IoT devices, as determined by their participation in bots, were Small Office/Home Office (SOHO) routers, IP cameras, DVRs, and CCTVs.
David Warburton, Senior EMEA Threat Research Evangelist, F5 Networks, said: “IoT devices already outnumber people and are multiplying at a rate that far outpaces global population growth. Increasingly, lax security control could endanger lives as, for example, cellular-connected IoT devices providing gateways to critical infrastructures are compromised.”
Distributed Denial of Service (DDoS) remains the most used attack method. However, attackers in 2018 began adapting Thingbots under their control to encompass additional tactics including installing proxy servers to launch attacks from, crypto-jacking, installing Tor nodes and packet sniffers, DNS hijacks, credential collection, credential stuffing, and fraud trojans.
The most common method attackers used to discover and eventually infect IoT devices was through global internet scans looking for open remote administration services. Telnet and Secure Shell (SSH) protocols were the most popular, followed by Home Network Administration Protocols (HNAP), Universal Plug and Play protocols (UPnP), Simple Object Access Protocols (SOAP), and various other Transmission Control Protocol (TCP) ports used by IoT devices. Common vulnerabilities and exposures specific to IoT device manufacturers were also prominent routes to exploitation.
The report suggests a significant and growing concern that IoT infrastructures – the servers and databases to which IoT devices connect – are “just as vulnerable to authentication attacks via weak credentials as the IoT devices themselves.” F5 Labs has found that cellular IoT gateways are just as vulnerable as traditional wired and WiFi-based IoT devices. As many as 62pc of tested devices were vulnerable to remote access attacks exploiting weak vendor default credentials. These devices act as out-of-band networks, creating network back doors, and are widely dispersed, the firm says.
Notably there was a large spike in attack traffic in March 2018 that drove a 94pc decline in total Telnet attack volume from the first to second quarter of 1018. This is important, as the frequency of Telnet attacks typically tails off when cybercriminal shift their focus from reconnaissance scanning to targeted attacks aimed at building deployable Thingbots. The top 50 logged attacking IP addresses are all new. This represents a change from previous four reports where the same IP addresses consistently appeared. The report suggests a range of new threat actors are in play, or that existing disruptors are transitioning to new systems. Other new developments include the introduction of attacking IP addresses hailing from Iran and Iraq.
Most attacks still originate in the networks of telecom and Internet Service Providers that provide internet services to homes, small offices and larger enterprises. The trend has remained the same for 18 months and is expected to continue. Attackers typically rent systems in hosting centers to initiate the building of a botnet. Build efforts are then taken over by the infected IoT devices in telecom networks. The study found scant decrease in the global footprint of Mirai, which is the most powerful Thingbot yet to have launched an attack.
The number of Mirai scanner systems across the world dwindled slightly from December 2017 to June 2018. However, Europe remains the only region where Mirai scanner infections remained relatively static from December 2017 to June 2018.
Not only is the threat of the original bot still powerfully present, but there are also at least 10 Mirai offshoots to consider (Annie, Satori/Okiru, Persirai, Masuta, Pure Masuta, OMG, SORA, OWARI, Omni, and Wicked). Furthermore, Mirai’s step-siblings are capable of much more than launching DDoS attacks, and can deploy proxy servers, mine crypto-currencies and install other bots.
Warburton said: “We are stuck with over eight billion IoT devices around the world that, for the most part, prioritise access convenience over security. Organisations need to brace themselves for impact, because IoT attack opportunities are virtually endless and the process of building Thingbots is more widespread than ever. Unfortunately, it is going to take material loss of revenue for IoT device manufacturers, or significant costs incurred by organisations implementing these devices, before any meaningful security advances are achieved. Therefore, it is essential to have security controls in place that can detect bots and scale to the rate at which Thingbots attack. As ever, having bot defense at your application perimeter is crucial, as is a scalable DDoS solution.”
