Patents: Changing the World, Or Not?

Patents: Changing the World, Or Not?

I love reading classic patents. For the ones that are truly ground-breaking, there’s a feeling of “this is going to change the world”, and where there’s generally a better explanation of the methods than in academic papers.

We have had three highly successful spin-out companies (Zonefox, Symphonic and Cyan Forensics), and a new one finding its way (Memcrypt). Overall, we have always submitted patents, in order to protect our core IP, and also to support investment. I’ve been lucky that my university has an excellent approach to patenting IP, and often defend a patent, until a time that the spin-out company can buy it back at the cost of defending it. If you are interested in the patents I have been involved in:

• Resilient secret sharing cloud-based architecture for data vault. Expiry: 2036. Here.
• Digital Forensics (part of Zonefox innovation). Expiry: 2029 Here.
• Information Sharing (part of Symphonic innovation). Expiry: 2033. Here.
• Method for reducing false-positives for identification of digital content (part of Cyan innovation). Here.

An often critical advocate of the US patenting system is djb (Daniel J Berstein), and who has spent a good deal of his time criticising the patent system for areas of cryptography. For him, the patent system often suppresses innovation and can stop new areas from moving forward. This goes back to a truly ground-breaking innovation:

It was a patent that completely changed cryptography. Daniel outlines that Diffie and Hellman would have published the work, anyway, as an academic paper, and that the patent was an after-thought. In fact, Diffie and Hellman had already distributed preprints of their paper “New Directions in Cryptography” in 1976, while the patent was submitted more than a year later in 1977:

The US patent system states that a patent is invalid if it is disseminated more than a year after the publishing of a patent. Daniel believes that the patent should never have stood, especially as the Diffie-Hellman-Merkle patent was invalidated because an inventor had distributed six pre-print versions of the paper, 14 months before the patent was submitted.

What is interesting about Daniel’s critical analysis, is that many claims have made it through the system, even though there is prior art. One of these is Crandall’s patent related to elliptic curve cryptography and in submitted in September 1991:

But, Dan outlines that Neal Koblitz — the independent co-creator of elliptic curve cryptography (ECC) —found some prior art that invalidated the patent, and which was published within the CRYPO’ 89 conference:

Another classic patent created by Brickell, Gordon and McCurley, focused on the way that we perform the exponentiation function (M^e mod p) and which is used within RSA:

But, again Dan disagreed with the uniqueness of the claims, and that it had been published by Yao in 1976, and with updates by Knuth in 1981.

Some great patents

Well, if you want to read more, try here. Dan has recently been critical of the usage of patents with PQC (Post Quantum Cryptography), and that the methods that will be adopted should be patent-free.

Anyway, here’s my six favourite patents that laid the foundations of Cybersecurity:

and the patents that have enabled Internet security:

Conclusions

And so? Well, we have always found that investors like patents, so if you are raising investments, they should be considered. But, if you want your method to being used and create impact, perhaps publishing the paper in a top journal is a good approach. Remember, too, that you should try and limit the dissemination of your new innovation before submitting the patent, otherwise it could be rejected.

So, … make 2021 the year of innovation … and let’s change the world …