The transition to new NIST standards for post-quantum cryptography (PQC) is a leap forward in algorithmic security, but not the core architecture required to address the “harvest now, decrypt later” issue presented by quantum computers. Although we hope post-quantum cryptography is quantum-safe, we have no mathematical proof it is true. No one should consider an algorithm’s interoperability or backward compatibility when connecting mission-critical systems and highly sensitive proprietary data. While necessary for web browsers and other mass commodity applications, one-size-fits-all and single-point-of-failure solutions are intolerable for sensitive, high-value infrastructure. PQC-encrypted data will be captured with the complete confidence that it will be broken, as all previous generations have been.
Government advises “crypto-agile” implementation
Consequently, the government advises new PQC implementations must be “crypto-agile.” This means the new post-quantum cryptography algorithms must be easily replaceable if a flaw or weakness is discovered. There is only one PQC algorithm for key exchange, which is the essential core function of all secure PKI network communications. This begs the question: what do we replace it with? The stark reality is alarming: no Plan B algorithm for key exchange was standardized. The reason is because SIKE, the PQC alternate, was broken by a regular laptop two years ago, just before the final standardization phase. The emphatic confidence of a few that it won’t happen again is unacceptable with so much at risk to everyone.
Incremental improvements vs. strategic transformations
The problem is fundamental when the industry norm dictates the release of standards, best practices, and solutions that are both reactive and slow-to-develop. Meanwhile, attackers and techniques are evolving in real-time with powerful tools. There are many recent examples analogous to the post-quantum cryptography transition we face today. It is an incremental tactical improvement, rather than a strategic transformation, based on momentum earned over decades.
Learning from Zero Trust mistakes
Zero Trust is a newly popularized buzzword for the older concept of de-perimeterization and continuously proving identity. Oftentimes, this assumes the adversary is inside the network. The Solar Winds and Crowd Strike victims demonstrate the concept’s futility and deceptive connotations. Similarly, the Storm-0558 use of Microsoft’s MSA key to forge authentication tokens gave Chinese hackers extraordinary access to zero-trust compliant enterprises, including most US government agencies. This extends to hardware with Intel’s SGX vulnerability for purportedly impregnable chip-based cryptographic security. Lastly and most recently, there is the unpatchable encryption-breaking flaw in Apple M1, M2, and M3 chips. Re-defining “what we really meant” by these terms is moving the goalposts instead of developing more durable redundant security.
Adopting post-quantum cryptography is a critical first step
Poor implementations and defects in both software and hardware should not be a severe threat to cybersecurity. This is because it is sure to happen and will eventually be publicly exposed and exploited by many. However, durable security in depth is not another factor for authentication on a smartphone for an app on the same smartphone. This is obvious security theater, but redundancy is implied by the term MFA—literally accurate, but practically useless. It is almost unbelievable that in 2024, large enterprises will still tolerate the existence of so many single points of failure in cybersecurity. Adopting post-quantum cryptography is critical and must happen presently, but it amounts to an existential risk above all others. It is completely vendor-agnostic and affects every cyber tool that needs to communicate safely and respond to threats (see: all of them).
Quantum-era cryptography and cybersecurity must evolve
We should celebrate the arrival of new standards, but we must remember the transition to post-quantum cryptography does not assure quantum-era cybersecurity. PQC is purely classical in that regard, and utilizes the same architecture used for 1970s telecom networks. Our digital economy uses virtualization, containerization, the cloud, and other technologies to achieve vast redundancy. Quantum-era cryptography and cybersecurity must evolve to use them too. There is no reason to compromise and tolerate the implicit risk in any single point of failure.
In cybersecurity, “Si vis pacem, para bellum” must be replaced with “noli ferre ullum punctum deficiendi,” meaning “do not tolerate any single point of failure.”
