IoT Security Legislation – Recommended Parameters

Morey Haber, October 25th, 2016

IoT Security Legislation

My blog earlier in the week touched on standardizing security legislation for Internet of Things (IoT) devices, leading to a few conversations on what this could potentially look like. The massive Distributed Denial of Service (DDoS) attack on October 21, 2016 has raised some questions regarding the security risks these devices represent and what can potentially be done to stop the threat in the future.

Parameters for IoT Security Legislation

IoT devices themselves are not subject to traditional laws for recalls if the security posture cannot be mitigated. In other words, you cannot recall a DVR used in a cyber-attack like a defective cell phone or faulty toaster just because you cannot change the default password. With these characteristics in mind, the following parameters should be a part of any potential future cyber security legislation that governs IoT devices:

  • Internet connected devices should not ship with common default passwords
  • Default administrative passwords for each device should be randomized and unique per device
  • Changing of the default password is required before the device can be activated
  • The default password can only be restored by physically accessing the device
  • The devices cannot have any administrative backdoors or hidden accounts and passwords
  • The firmware (or the operating system) of the device must allow for updates
  • Critical security vulnerabilities identified on the device for at least three years after last date of manufacturer must be patched within 90 days of public disclosure
  • Devices that represent a security risk that are not mitigated or fail to meet the requirements above can be subject to a recall

While I fully expect the list to evolve, and that some of peers may object to entries or have some of their own (please send your ideas if you do), we cannot continue to allow unsecure devices to be connected to the Internet. This could jeopardize the infrastructure that we have all become so dependent upon.

While we must continue to invest in defensive technologies to stop threats already deployed, and threats from devices deployed abroad, if we do not act soon, the rapid adoption of insecure IoT devices could leave us with more potential attack vectors than actual legitimate devices. A simple combination of password management, vulnerability management and asset management can go a long way in starting this process.

Is it Time for the Government to Get Involved?

If you consider the potential address space of IPv6, and the potential adoption of devices from light bulbs to cameras that can be connected to the Internet, we need to adopt basic safe computing for all devices in order mitigate potential botnet threats from IoT devices like we experienced last week. The shear thought of our Internet infrastructure being disrupted by insecure devices being sold en mass from a foreign nation just raises more questions than and answers, and no current trade or legal methods to stop them.

After all, it has now been proven these commercial devices can be weaponized with malware and target the largest companies in the United States and cause millions in financial losses. It is time for our government to step in and mandate the basics.

What are your thoughts? Let’s keep this important conversation going.