The Internet of Things and the Inevitable Collision with Products Liability PART 2: One Step Closer
The Internet of Things and the Inevitable Collision with Products Liability, published in February 2015, identified a number of factors leading to the emergence and phenomenal growth of the Internet of Things (IoT). It also identified issues relating to potential product liability exposures and the impact that IoT-connected devices could have on risk assessment and risk transfer due to the consequences of foreseeable vulnerabilities and failures with IoT-connected products.
This second article addresses in more detail the emerging liability risks for the stakeholders at the forefront of the development and implementation of these technologies who, in turn, will be forced to confront those liabilities whether or not they are prepared to do so.
Several documented IoT failures have already occurred in 2015. Notably, Wink’s wireless hub, which is connected to a variety of devices in homes via a single app, experienced a failure in April that disabled the connected devices, potentially leaving consumers vulnerable to breach of their home security systems or other failures. Chamberlain and Ooma also experienced failures, both of which involved compromised IoT connective services and resulted in disruptions that had the potential to affect customers’ home security.
The first IoT class action was brought in March 2015 spurred by the February publication of a report by U.S. Senator Edward Markey (D-MA), which was also covered on a CBS broadcast of 60 Minutes. The action was brought against Toyota Motor Corporation, Ford Motor Company and General Motors LLC. (See Cahen, et al. v. Toyota Motor Corporation, et al., U.S. District Court of Northern California, San Francisco Division, Civil Action No. 4:2015cv01104.)
Senator Markey’s staff questioned 16 auto manufacturers regarding how they protect against vulnerabilities of vehicles to the threat posed by outside hackers infiltrating vehicle systems that could lead to loss of control over vehicles or disabling of safety devices. (See Tracking & Hacking: Security & Privacy Gaps Put American Drivers at Risk.)
The investigation was prompted by studies that disclosed that hackers can get into the controls of some popular vehicles, causing sudden acceleration, turns, loss of brakes, activation of the horn, faulty operation of the headlights, and modification of the speedometer and gas gauge readings.
Senator Markey’s investigation was therefore directed at determining what automobile manufacturers are doing to address these issues and protect drivers.
According to the report, based on the auto manufacturers’ responses, a number of serious vulnerabilities were identified:
Nearly 100 percent of cars on the market include wireless technologies that could pose vulnerabilities to hacking or privacy intrusions.
Most automobile manufacturers were unaware of or unable to report on past hacking incidents.
Security measures to prevent remote access to vehicle electronics are inconsistent and haphazard across all automobile manufacturers.
Only two automobile manufacturers were able to describe any capabilities to diagnose or meaningfully respond to an infiltration in real time, and most say they rely on technologies that cannot be used for this purpose at all.
Additional concerns about driver privacy were identified as navigation systems and other features can record and send location or driving history information. This topic will be explored in greater detail in a future segment of this series.
The complaint filed in the resulting class action brought against the auto manufacturers closely mirrors the threats identified in the Markey report. The core allegations made against each auto manufacturer are based on breach of warranty claims that the vehicles are not free of defects: “Because defendants failed to ensure basic electronic security of their vehicles; anyone can hack into them, take control of the basic functions of the vehicle, and thereby endanger the safety of the driver and others.”
The complaint further alleges that each vehicle has up to 35 separate electronic control units (ECUs) that interact with controlled area networks (CANs) and “vehicle functionality and safety depend on the functions of these small computers, the most essential of which is how they communicate with one another.” As a result, a “hacker could take control of such basic functions of the vehicle as braking, steering and acceleration – and the driver of the vehicle would not be able to regain control.”
This action is still in its early procedural stage, so forecasting the merits and outcome is premature, but nonetheless it should provide cause for concern for manufacturers and software companies that are actively developing products for the IoT marketplace.
Other Threats Identified
In April 2015, the U.S. Government Accountability Office (GAO) issued a report addressing commercial aircraft safety from cyber threats. GAO noted that “modern aircraft are increasingly connected to the internet, [but this] interconnectedness can potentially provide unauthorized remote access to aircraft avionics systems.” (See Air Traffic Control – FAA Needs a More Comprehensive Approach to Address Cybersecurity as Agency Transitions to NextGen.)
Among GAO’s conclusions, the report found “…FAA has taken steps to protect its ATC (Air Traffic Control) systems from cyber based threats; however significant security control weaknesses remain to ensure the safe and uninterrupted operation of the national airspace system.”
The FBI is reported to be investigating an individual who claimed through social media that he had hacked into passenger airplane controls while on board flights and had taken over command of certain airplane functions. The intrusion was reportedly made through Wi-Fi access via the plane entertainment system. (See The Washington Post, May 18, 2015.)
Swiss Re in May 2015 published a global risk assessment report in which it identified the Internet of Things as among the highest potential risk impact comparable only to de-globalization, the great monetary experiment and supernatural category storms. (See Swiss Re SONAR, New Emerging Risk Insights.)
More recently, AIG has published part one of a series of white papers addressing IoT risks, The Internet of Things: Evolution or Revolution. The report predicts significant risks for businesses entering the global market for IoT-connected products:
“From cyber breaches to shifting questions of property and products liability, businesses cannot afford to enter this new technological world unprepared. For example, every object that connects with the Internet is another entry point through which the cyber-criminals can enter a business’ [sic] enterprise system. Equally dangerous, in a world where machines replace humans as the decision-makers and sensors are continually capturing data, serious questions of liability, resulting physical damage and privacy arise.”
As to liability concerns, the paper posits a number of thought-provoking scenarios:
“When it comes to autonomous vehicles, like driverless cars, we are faced with an obvious ethical dilemma: In the seconds before an accident, should an autonomous vehicle do anything it can to protect the passengers, even if it means harming other motorists or pedestrians? When humans are behind the wheel, collateral damage, as terrible as it is, doesn’t pose much of an ethical problem. A human being in danger can’t be faulted when its survival instincts make it swerve its car into a pedestrian. But when machines are the decision-makers, does a pedestrian harmed in accident have a case against the car manufacturer? Does a driver have a case against a car manufacturer following an accident in which he or she was injured?
“IoT devices also raise troubling questions when it comes to device malfunction. Sensors can be embedded in critical infrastructure like dams, bridges, and roadways to monitor structural integrity as well as environmental conditions that could undermine structural integrity. A road near a flood area could be embedded with sensors that know the moment rainfall has exceeded a point that gives engineers advanced warning of flooding. Indeed, protecting infrastructure is one of the most exciting aspects of IoT. Yet when we turn more and more of our critical infrastructure and security systems over to IoT objects, we run the risk of a catastrophe if and when those objects fail. We can apply this to the private sector as well. To cite a non-lethal example, in April 2015 several American Airlines flights were delayed when a software malfunction rendered pilots’ tablets, which they use for navigational purposes, useless. Although the malfunction was easily fixed with a software update, these examples show just how exposed we already are because of our connected devices. When—not if-- they fail, will we be prepared?”
Supply Chain Considerations
Software vendors and sensor manufacturers are now critical component part suppliers in the development of IoT-connected products. Major players in the development of IoT products and applications are acquiring software companies and partnering with Internet start-ups to take strategic advantage of the emerging market. Google, Microsoft, Samsung and Amazon have all made recent acquisitions of companies that will accelerate their penetration of the multibillion-dollar IoT marketplace. Strategic joint venturing between technology companies and other business enterprises seeking to catch the wave are also occurring weekly. These new strategic partnering initiatives will have an impact on component part suppliers’ product liability exposures, most notably the software vendors and sensor manufacturers.
Component parts manufacturers have long been subject to product liability exposures when a critical component part causes or fails to prevent a product failure.
Sensor manufacturers will face greater liability exposure in part simply because of the greater use of sensors in all manner of IoT product applications. Sensors have already been the subject of product liability claims and lawsuits for alleged failures in products such as smoke detectors, carbon monoxide alarms and automobile airbag systems.
In the IoT world, software licensors will not be protected against third-party injury claims. Many software vendors have either (1) been unaware of their product liability exposure to claims and lawsuits for bodily injury and property damage caused to third parties or (2) have failed to provide for such exposures in their agreements. Licensing agreements and their built-in provisions for protection against failures have largely been limited to instances of failures or damages between a software vendor and its customer, and specifically related to the task for which the software was provided. These agreements may not insulate a software vendor from liability resulting from a failure that injures a third party or causes property damage to a third party for which such loss was foreseeable.
Insurance coverage for losses that result in property damage or bodily injury is an area of vulnerability for stakeholders. Traditional cyber-data breach insurance coverage addresses the loss in intangible property as a result of a breach. An IoT product failure that results in property damage or bodily injury will, in the absence of a specifically designed policy, require companies to look to the traditional coverage afforded under CGL, PL, D&O and E&O policies. Inevitably, in the absence of specifically designed coverage there will be instances where gaps will exist and the exposure will be uninsured.
Software companies and product manufacturers need to and will develop contractual language to properly balance and shift the potential third-party liability exposures. However, the enormous financial power differentiations between the big technology players and the medium-sized to small software start-ups and new IoT-centric businesses will require that adequate financial protections in the form of IoT insurance coverage be developed. The insurance industry to date has been slow to recognize the enormous potential market for IoT insurance coverage for commercial liability exposures. It has been predicted that companies such as Google, Amazon and Apple with their huge liquidity may jump in and take a leading role at the expense of insurers. This development has already begun to unfold in personal lines insurance where technology companies have a huge advantage over insurers with the collection and use of data from consumers. Insurers are also coming around to recognize the power of big data and have initiated their own strategic partnering with technology companies such as American Family Insurance Company and Microsoft’s joint enterprise to create an accelerator for startups focused on smart home technology.
Lack of Standards
One of the immediate short-term concerns for stakeholders is the lack of uniform standards for the IoT whether it is industrial IoT, consumer goods or other applications. The lack of uniform standards will result in vulnerabilities for IoT companies when the inevitable accidents occur leading to claims and lawsuits. Plaintiff attorneys will be certain to seize on the lack of self-governance within the industry based on the lack of recognized minimum standards for safety and security.
Currently, standards-setting organizations are working on developing standards that will be implemented at some point in the future. These include:
Institute of Electrical and Electronics Engineers (IEEE): P2413 Draft Standard for an Architectural Framework for The Internet of Things Working Group
International Telecommunications Union (ITU): Y2060 – Overview of The Internet of Things.
Industry security specialists have sounded the alarm over concerns with the fast-paced development of IoT without adequate security safeguards. A June 14, 2015, interview of two security specialists that appeared in The Globe and Mail noted: “The key weakness of most tech companies and their Internet of Things (IoT) customers is a failure to create a ‘threat model’ and test security against that. If they don’t know what they are trying to defend, and who they are trying to defend it against, any security measure and no security measure applies. You attack the weakest device, and an IoT device usually has weak or no authentication with other devices in the same network.”
Another security specialist noted in a Pace interview:
“Despite the industry’s best efforts, the IoT will never be 100% secure … We don’t know all of the ways that smart devices will interact with each other and how they will be used. The complexity and scale of the IoT will inevitably lead to security holes. A detect-and-respond mindset must be adopted from the start. … Manufacturers and other businesses should assume that the IoT technology stack will be attacked, and be properly prepared to respond. This means investing in systems that automate the detection of malicious activity so that it can be contained and remediated before data is lost or damage is done.
“Users can’t be expected to download antivirus software for every smart connected device – it may not even be possible given the disparity of operating systems. At the same time, businesses can’t be expected to deploy patches and updates to disposable, lightweight devices. IoT devices must be built with security and privacy controls baked in. Networks must be instrumented to automatically detect malicious behaviour.”
Next up in Part 3. The IoT and privacy as it relates to data collection from IoT devices. Who owns the data? Who is responsible for its security? What steps are necessary to inform and protect consumers’ data from unauthorized uses or hacking threats? Also, what will be the reporting obligations for IoT product defects to government safety agencies? Who will be obligated to report? What event may trigger an obligation to report when there is a threat of physical damage or bodily injury arising from an IoT device defect?