We dance round in a ring and suppose,
But the Secret sits in the middle and knows.
– Robert Frost
How can you trust a secret if you can’t tell where it came from? Today, we are excited to share that the Internet Society is a proud supporter of the Cryptech initiative. From the website introduction:
Recent revelations have called into question the integrity of some of the implementations of basic cryptographic functions and devices used to secure communications on the Internet. There are serious questions about algorithms and about implementations of those algorithms in software and particularly hardware.
We are therefore embarking on development of an open hardware cryptographic engine that meets the needs of high assurance Internet infrastructure systems that rely on cryptography. The open hardware cryptographic engine will be of general use to the wider Internet community, covering needs such as secure email, web, DNS, PKIs, etc.
In his recent blog post on Pervasive Internet Surveillance, Phil Roberts enumerated some of the technical efforts underway in the IETF as a result of this past year’s ongoing revelations about unwarranted surveillance and data collection. The IETF is not the only organization involved in attempts to block or mitigate what some have characterized as an attack on the Internet itself. Recent announcements include attempts to secure email (ProtonMail, Darkmail, etc.), to secure your mobile phone (Blackphone), to put better tools in the hands of end users (leap.se), and to provide private messaging services (jabber/otr), to name just a few. There are also serious efforts to secure the routing and domain name infrastructure, which rely on underlying methods for generating and protecting strong keys for encryption.
All of these projects depend on strong encryption and strong keys as tools to both establish trust and to protect sensitive data. Sadly, trust in many of the components needed to achieve real secrecy has been undermined. Core components such as entropy, cryptographic algorithms, and purpose-built hardware have all been compromised. Some of this has been malicious, some was the result of pressure from governments, some was due to the failure of business practices in deployment, and some was the result of underfunded and understaffed volunteer efforts in maintaining widely used software. The combined result has lead to widespread doubt in formerly trusted institutions, developers, and vendors. For many, even those using these products and tools as part of their critical infrastructure, the process of key generation and key management is a black box.
Here is the paradox: how can you trust the generated secrets if you can’t tell how the box was made? The Cryptech project poses the following riddle: when do you want your black box to be transparent? The answer: while you’re building and configuring your own particular box.
The Cryptech effort began in late 2013 with a small group of engineers at a side meeting at IETF 88 in Vancouver. The project has strong support from the IETF and IAB chairs but the project is not limited to IETF participation. While early use cases included IETF protocols such as RPKI and DNSSEC, there was also interest from Certificate Authorities, the TOR Project, and others. Cryptech is aimed at those processes requiring a very high degree of assurance – normally provided by purchasing a Hardware Security Module (HSM) – but in this case they will replace the closed box with an open one. One of the project objectives says: “The intent is that the resulting open hardware cryptographic engine can be built by anyone from public hardware specifications and open-source firmware. Anyone can then operate it without fees of any kind.”
Stephen Farrell, an IETF Security Area Director, had this to say in supporting the project:
“The particular aspect that the cryptech folks are addressing is that there is somewhat of a crisis of confidence in the implementation of the cryptographic functions that underpin our most important Internet protocols. These functions are required for securing the web and for many other aspects of Internet infrastructure such as DNS security and routing security. Cryptech and the team that has been assembled could significantly help to alleviate these specific concerns by producing open-source hardware designs that can be directly used or re-implemented by others. A significant benefit of that is to provide confidence that the design and implementation is as free from potential nation-state or other interference as can be. While it may never be possible to achieve 100% confidence in that, it is definitely technically possible (though non-trivial) to do far better than we have to date – today we essentially have a choice between pure software cryptography or commercial hardware products for which its impossible to see what’s “under the hood.””
The Internet Society is in strong agreement with this statement and has been an early supporter of this effort. We are actively encouraging additional partners. The project will need steady funding sources, active coders, good community review, and additional experts to audit and validate the project outputs.
The Cryptech team has been evangelizing the project and will be on hand at IETF 90 for a project update. The briefing will be held on Wednesday, 23 July, during the lunch hour in the Quebec room of the Fairmont Royal York Hotel in Toronto. Please come and bring your questions and your inputs.
Details of the Cryptech project can be found here: https://wiki.cryptech.is. The project is hosted by SUNET, but participants come from a variety of sources including academia, the open hardware and software communities, the TOR project, and the IETF. The team is committed to a transparent development process and early code is available at https://cryptech.is/browser.
Please join us in this effort to help make the Internet a safer place.