The Internet of Things (IoT) has been a driving force of digital transformation; as such, IoT devices' notorious lack of proper security is stunting both IT security and IT growth. In the first segment of this series, we began to explore the difficulties that IT departments face in digital-transformation initiatives where cloud migrations are concerned. One of the reasons cloud is vital to digital transformation is the enablement of IoT, bearing the load of exploding numbers of new devices of all kinds on the network. (See Digital Transformation With Cloud: Answering Risks With Algorithms.)
Moreover, by their very nature, many IoT devices when first conceived were typically never intended to offer connectivity or contend with cybersecurity issues -- and IoT manufacturers haven't exactly done much to enhance device security at nearly the same rate as they have connectivity.
Rules and standards
"The gotcha with any IoT deployment is how those devices are kept secure over time," Chris Smith, head of global security services at CenturyLink, told Security Now. "We still see a general lack of accountability with manufacturers when it comes to patching and firmware."
A legally mandated floor -- albeit a very low and somewhat amorphous one -- will be coming to the US in 2020, when a California law impacting all devices with connectivity will come into effect. The law will require device-makers to equip devices sold in California with "reasonable" and "appropriate" security features -- such as unique default passwords, or compulsory user-generated authentication on first use. (See California Looks to Pass Rudimentary IoT Security Legislation.)
There are still yet bigger problems to address in the connected-device industry, like that of customized protocols in lieu of standardization -- which, by their one-off and usually proprietary nature, do not lend themselves well to security research.
"One of the hard things with IoT security is that the full architecture is not yet standardized," Ben Stori, a consultant at Solution Design Group, told Security Now. "Every vendor has their own architecture to run their devices on."
Last week, the European Telecommunications Standards Institute (ETSI) Technical Committee on Cybersecurity (TC CYBER) announced its release of TS 103 645 -- which purports to be the "first globally applicable standard for consumer IoT security ... provid[ing] a basis for future IoT certification schemes." According to ETSI, compliance with TS 103 645 would prevent many headline-grabbing IoT exploits -- such as those used to spy on the users -- from being as successful in the future.
"The tools to tighten [IoT] up have to be applied in the original design; this is the whole ethos of 'secure by default'," Scott Cadzow, a security expert from TC CYBER, told Security Now. "[The] key to 'secure by default' is understanding and working out where an adversary will try and subvert the system."
Securing endpoints inside instead of outside
Perimeter defense, for instance, has become so porous as to make perimeters almost not worth defending (almost
; see below) -- digital transformation or no. For this reason, SAP global CSO Justin Somaini proposes instead making application security the first line of defense -- to the point of developing "self-defending apps
" with the ability to regulate and restrict access and modification on their own in real time.
Meanwhile, recent studies and reports have shown that even though endpoint attacks are on the rise, actual physical compromises of endpoints remain a rarity. More to the digitally transformative point, the complexities and obstacles to securing certain IoT endpoints can be plenty alleviated with strong application security -- especially when buttressed by solid perimeter defenses. (See Data Breach Increase Shows Endpoints Are Under Attack and Should All IAM Be CIAM?)
Seeing is securing
This is not to suggest ignoring the physical security of IoT devices, for they can be compromised via physical access in a number of ways. Network visibility thereby becomes all the more important -- though it is difficult to track IoT devices and assess their respective threats and vulnerabilities as they exponentially widen the attack surface.
"Log monitoring can help, but it must be able to accommodate all the sources the business has," said Smith. "And it needs to support not just threat detection but [also] the appropriate level of analysis to avoid putting out a flood of false positives."
High false-positive rates represent an inherent weakness of standard threat-intelligence tools; they lead to security-alert fatigue and IT-operations slowdowns -- defeating the whole point of digital transformation's agility goals in the first place. Cybersecurity-startup advisor Chris Richter told Security Now that many security frameworks even just a couple of years old already need to be revisited and updated to better pinpoint modern IoT-security threats -- recommending the NIST Cybersecurity Framework as a starting point in the absence of something better. (See Endpoint Security: 3 Big Obstacles to Overcome.)
"[Enterprises] are embracing IoT at an alarming rate," said Richter. "Any digital-transformation initaitve must include a cybersecurity-transformation element as part of the plan to begin with. Otherwise, it's very difficult for cybersecurity leaders and CISOs to stay current."
'There are lots of simple mitigations," adds Cadzow. "[But] there is no such thing as a benign device attached to the Internet."
—Joe Stanganelli is managing director at research and consulting firm Blackwood King LC. In addition to being an attorney and consultant, he has spent several years analyzing and writing about business and technology trends. Follow him on Twitter at @JoeStanganelli.