A new father’s musings on the problems with securing embedded systems, and why there are so few incentives for architecting trustworthy IoT technology from the ground up.
This weekend, I hit a new milestone. I celebrated my first Father’s Day as the parent of a week-old newborn. Like any new father, I’m a mix of pride, exhaustion, and unnecessary amounts of caffeine and ibuprofen. Unlike some new fathers, I spend an exorbitant amount of time thinking about Internet of Things security, and how critical security it is to the future of not only embedded systems, but our society.
My concern drove me to become one of the first researchers to join the Internet of Things movement back in 2010. In 2011, I successfully performed the first remote hack of a car, demonstrated at the Black Hat Briefings in Las Vegas. Many other research projects followed. More importantly, I won a DARPA grant in 2012 focusing on IoT security (then dubbed Machine to Machine) and created a threat model for embedded systems, IoT networking technology, and the back-end services that drive them. Due to a surprising lack of interest in IoT security at the time, I had to wait to release my findings until the ecosystem gained momentum. After some negotiation, I teamed up with the GSMA in early 2015 to transition my DARPA research to the GSMA IoT Security Guidelines project, which is now publicly available for download on the GSMA website.
Yet, years of research didn’t prepare me for the moment of clarity that had, presumably, been brewing in my head for quite some time. It took my baby boy, Pierce, to put the pieces together for me. I’ve been saying it for over five years, but the Internet of Things is not simply a fad or a buzzword. The IoT represents the next generation of computing — the hybridization of our physical and digital worlds. As our physical world becomes more interconnected with our digital world, identity becomes the center of not only computer security, but computing, period.
And this brings up a fascinating yet often debated point regarding consumer and industrial systems: How can these devices be allowed to control our physical lives when we simply can’t engineer trustworthy technology?
That’s when it hit me. Raising a child requires sacrifice. It requires discipline, patience, and most of all, it requires a safe environment for the child to grow. Children require education and, possibly more importantly, the opportunity to play, in order to strengthen their intellect and social skills. But where does education and the right to play come from? Safe and secure environments. Children must be provided with a safe and secure environment for them to explore, learn, and socialize. Whether that environment is a country, a town, or a school, children that live in volatile or dangerous environments have less ability to focus, fewer opportunities for meaningful work, and more stressful social relationships. In essence, a society’s future is founded upon the safety and security offered to its children.
Truly, technology is no different. We’ve built processing architectures and platforms in an attempt to create safe and stable environments for researchers and entrepreneurs to grow incredible children of their own. From cancer research, to the Large Hadron Collider (LHC), to Sexual Health Innovations open-source Callisto Project, our society will benefit from these conceptual children, birthed from evolving computing environments.
And yet, we are failing our technological children. Internet of Things technology represents a massive influx of embedded systems into not only the Internet, but our global society. The platforms these devices are designed from are riddled with security flaws. IoT security has become such a punch-line in the information security community that Dave Aitel coined the phrase “junk hacking” to describe how easy it is. The reality? He’s right.
Modern embedded security is a problem. There are few, if any, incentives to architect secure technology from the ground up. The incentive is, instead, to build and deploy a Minimum Viable Product (MVP) as quickly and cost-effectively as possible. Why? Simple. This equates to investments and tangible sales. There is no stop-gap to accentuate the requirement of security, even though designing secure embedded systems can literally save a startup millions of dollars in future engineering, device replacement, insurance, and legal costs. The incentive has to change.
Recently, the Federal Trade Commission (FTC) evaluated the IoT and concluded that security and consumer privacy are rapidly growing concerns. Yet, they have stated in February of this year that the FTC does not intend to enforce regulation on the IoT. If our lawmakers and regulators don’t stand for consumer privacy and safety, corporations that do not yet understand the risks they are putting their own businesses in will see little reason to invest in architectural changes.
The result, as we already know, will be disastrous. You may not have to worry about physically getting your pocket picked in Amsterdam, but your rental car’s telematic system may be a backdoor to the NFC payment application on your smartphone.
As a community, we need to step back and take a moment to reassess how we raise our children in the technology that is quickly evolving in the Internet of Things. Do we enforce regulations and laws that ensure they will grow on a platform that incentivizes privacy, confidentiality, and integrity? Or will we raise them with promiscuous interfaces, loosely authenticated APIs, and open doors to criminal behaviors?
Don A. Bailey is a pioneer in security for mobile technology, the Internet of Things, and embedded systems. He has a long history of ground-breaking research, protecting mobile users from worldwide tracking systems, securing automobiles from remote attack, and mitigating … View Full Bio