On the Road to a Safe and Secure Internet of Things: What Companies Should Do
Originally appeared in Kaye Scholer’s Fall 2016 Consumer Products: Adapting to Innovation Report.
This article was the most read legal thought piece in the US in September 2016 (Mondaq).
—By Adam Golodner and Paul Margulies
The growth of the Internet of Things (IoT) seems hard to overestimate. Cisco predicts that by the year 2020, there will be 50 billion devices connected to the internet, and some of those will be connected to other IoT devices as well. IoT touches consumer home and wellness products, automated driving, smart manufacturing, smart energy, smart farming, and earth and ocean observation systems, to name some prominent examples. IoT devices, sensors and cyber-physical elements will transform companies and industries, driving efficiency and creating new services throughout the world. Given the benefits, and low cost of computing and connectivity, this march forward is unstoppable.
What does this mean for companies and their customers? Big changes, most all good. But it does require that companies understand and plan for them now. For most companies, the transition to the IoT means that they are now or soon will be an IoT company. If you put software and connectivity in your product, or use it in the elements of your manufacturing, supply chain or distribution, you are an IoT company. Have you thought about what that means to your internal and external business processes? And have you gamed out what you must do to ensure you are a safe and secure IoT business partner? Specifically, how can you make sure your product has inherent integrity and what’s your plan to update or patch it, secure it, manage its communications, ensure customer privacy and control its interactions once it’s connected in the field? And, from a legal perspective, how have you managed any downside risk?
Reasonable security experts are concerned that we are putting IoT elements out into the infrastructure, without enough security wrapped around the devices or their connectivity. They worry that in the rush to get IoT out there—which market and consumer demand is driving— we are “making the same mistake we did with the Internet” of not baking in security from the start, but merely bolting it on later. The concern centers around: the “low security” IoT element becoming a vector into the connected home network or enterprise network, allowing easy access into otherwise better protected elements (soft underbelly); the IoT function itself being easily disrupted, halted, manipulated, or corrupted; the possibility of redirecting IoT data streams or functions; the volume of “private” information being collected, created and transmitted; and the potential to take over significant numbers of IoT elements to launch DDoS attacks. All of these are thinkable, and most of these are things companies can recognize and plan against.
An Enterprise and Product Issue
For companies that produce products, IoT is both an enterprise and product issue. From an enterprise perspective, if you are providing your customer with connectivity, you have a role in the consequences of that fact. Just like information technology (IT) companies that sell software and hardware, you have to plan for the expected bugs in the product, figure out how you are going to patch, update, and respond to reports of vulnerabilities, and make sure you are not putting an attack vector in your customer’s home or business. By connecting your previously unconnected product you have just become an IT company—congratulations! Now you have to act like one too. Assign an owner, create product integrity teams, use a software development lifecycle process, test or certify products before shipping, create a product-incident response team, determine your vulnerability disclosure process, determine how to patch or update products, and get ready for constant maintenance, response and exercise for crisis control. There are models in the IT industry to follow, and you now need to be able to move in internet speed to ensure a safe and secure ecosystem.
You Are Part of an Ecosystem
Your device is now part of an ecosystem. You have to decide how your product is going to connect to other elements in a network, a home, an enterprise, the internet, the cloud, the world wide web. IoT devices are different from most other ecosystem parts. They are low power, light hardware and slight software, yet we want them to be fully safe and secure. Part of the challenge today is that IoT technical standards are just now being created, which means we are still in a standards war, globally, so you have to make some choices. Are you going to use and align with IETF or IEEE or ISO or US or Chinese or other standards? There are also private industry consortia creating standards, such as the Industrial Internet Consortium and the Alliance for Internet of Things Innovation. You have to get involved and make some bets.
You also have to decide how you are going to help ensure security of the devices and privacy of the information. How will you or the user determine if there is a security issue and how will you manage that issue? Is the home or enterprise network going to manage security through intermediate device anomaly detection, use a firewall at the gateway, or use another device for the enforcement of security policy, and how is that accomplished on the IoT device or data stream? Are you going to manage the device through your own private cloud? Public cloud? Who owns and controls the cloud infrastructure? And then how do you control the IoT elements? Who “owns” the data, who can access the data, created by the IoT device (the user, you, a third party)? Can and should the communications from the device be “whitelisted,” only allowed to communicate with you, or limited service providers, or select geographies? How do you control data streams? Are machine-to-machine interactions allowed, or controlled? Do you need to have human intervention before machines autonomously do things to each other and the infrastructure? All these issues need to be understood and decided before putting the device into the infrastructure. Each of these decisions has legal, security and privacy consequences.
Legal and Regulatory Issues
The IoT creates a host of new legal and regulatory issues. On the legal side, the general principles of tort and contract apply, but with a few new things to think about. In the past, tort liability for issues related to software has been quite limited. Plaintiffs generally could not recover in tort as injuries relating to software tended not to result in physical injury, and were therefore precluded by application of the “economic loss rule.” In the IoT, however, a software issue (with autos, medical devices, airplanes, construction tools, HVAC) may result in physical injury and a new line of tort liability. In 2013, for example, Toyota was subject to a $3 million wrongful death verdict due to flawed throttle control software. The issues become even more complicated when an external event like a hack occurs and liability needs to be established.
On the regulatory side, although there are no new IoT specific regulations (yet!), policy makers are very concerned about the impact of IoT. In the US and EU, the same privacy rules that apply to data in general also apply to the IoT. Every company will have to do an IoT compliance review to see where their new IoT business intersects with existing privacy and security laws. Moreover, in the EU, the EC has set out a number of nonbinding considerations for the IoT, such as product safety, allocating liability, standardization frameworks and possible product certification requirements, and the EU Article 29 Working Group of data protection commissioners has set out some as well, including data privacy by design and ensuring user consent for data collection. Many sector regulators—autos, airlines, medical devices, electric—are working through whether and what voluntary or mandatory security, safety and privacy requirements may be needed for the IoT. This is a process all IoT companies should engage in with their governments, as the rules are being written now and can affect business models profoundly.
C-Level and Board
After determining the extent your company is an IoT company, and working through the issues and decisions set out above, it’s time to talk with company leadership and the board about what it means to be an IoT company, the new processes the company may have to put in place, the choices about how to engage with the ecosystem, and how to manage the IoT risk. Very often, as companies move into the IoT driven by advances in technology and competitive forces, there is not time to pause, see how their relationship to their customer and the network has changed, and make structured decisions about how they want to represent themselves in this new IoT world. Outside of pure tech, the board of directors often has not spent time understanding what IoT means, whether the company is in fact an “IoT company,” and what it means to internal processes, competitive advantage, and business and regulatory risk. This is a conversation that needs to happen, and one the board will appreciate. It’s an exercise in translating technology to business impact and, although it can be challenging, it is better to bring the board along at the beginning of the transition, instead of the end, when some problem hits and brings unwanted media attention.
Congratulations! You are now an IT company. As you continue along this journey, you need to reexamine your relationship to your customers, the data that will be created, the processes you need to put in place to ensure a safe and secure environment at internet speed, and the new legal and regulatory issues and risks that need to be managed. By taking the time now, the risks can be managed, and you will have lots of good things to talk about with the board as you introduce them to the Internet of Things.