Encryption patentable in the US: Is this the rise of US software patents?


Encryption of data is vital in many computing systems and has recently been found to be patentable in the US, at least if the patent is specific enough about the encryption algorithm. If encryption is patentable, then many other software and computer implemented inventions should also be patent-eligible in the US.

Encryption involves abstract manipulation of data and so the patentability of encryption in a computing system is an interesting question, in light of recent decisions like the US Supreme Court's decision Alice [1].

In the Alice decision, a two-step test for patent-eligible subject matter was created. In a first step, it is considered if the claims are directed to an abstract idea. If so, in a second step it is considered whether the claims have a sufficient "inventive concept" to render the idea patent-eligible. A summary of the two-step test can be found here.

I do not believe that anyone would dispute that encryption is directed to an abstract idea. In fact, many computer implemented calculations relate to an abstract idea and so this question is particularly interesting.

The US District Court (Eastern District of New York) recently published the decision PAONE v. MICROCHIP TECHNOLOGY INC. [2]. In that decision, Justice Cogan considered the patentability of a computer implemented encryption system, which is the subject of US patent No. 6,259,789 by PAONE.

What immediately stands out is the generic nature of the computer system features recited in the claims of the patent. It appears that his Honour considered, at least in part, that the specificity of the encryption algorithm itself raised the invention above a mere abstract idea, in the context of communicating computers. This conclusion may be somewhat surprising to some, and may indicate a significant softening by at least this US court in considering what constitutes patent eligible subject matter in the area of software patents and computer implemented inventions generally.

Claim 23 is a characteristic claim of the patent, and is:

A cryptographic communications system comprising:
at least two networked computer systems linked by a communication channel; and

each computer system including a central processing unit and a memory storage device for executing a block cipher encryption/decryption process;

wherein the encryption process transforms an input plaintext message to a ciphertext message and the decryption process transforms the ciphertext message to the input plaintext message, the encryption/decryption process using at least one dynamic object key which is modified using a non-linear function for each block of input data, each object key being associated with a different key schedule to encrypt/decrypt the input plaintext/output ciphertext message.

The generic computer system features have been underlined by me. The “inventive concept” is the non underlined portion of the claims and is, in my view, rather abstract. His Honour characterises the invention as computer-improved methods of encrypting data, as well as certain cryptographic communication systems that employ the same methods, and which use a particular block cipher.
His honour focused on the specific nature of the encryption algorithm considered, and was of the view that the limitations of the specific encryption algorithm very much diminished the pre-emptive effect of the claim. 

While His honour was of the opinion that while software patents may have been called into question following Alice, a patent for a method of data encryption is not per se invalid, as long as it is specific enough. His Honour was further of the view that while encryption in general represents a basic building block of human ingenuity that has been used for hundreds, if not thousands of years, the patent does not claim process that can or does involve the encryption of data for some purpose that is otherwise abstract. Rather, it climbs a specific method of doing so.

This is at least the second US decision where an abstract idea performed on a computer has been found to be patentable and suggests that US software patents are not unobtainable, at least for some inventions.  Those interested may wish to review my post on the VERACODE decision [3], in which it was found that a software decompilation method was patent eligible in the US.

In conclusion, it appears that a patent for software or other computer implemented invention that is directed to an “abstract idea” is more likely to be valid if the claims are drafted to a specific solution to a problem, rather than having a broad scope.

[1] Alice Corp. Pty. Ltd. v. CLS Bank Int’l, _U.S._, 134 S. Ct. 2347 (2014)
[2] Luciano F. Panone - against - Microchip Technologies Inc., Memorandum Decision & Order NYED-2-15-cv-00596 (2015)
[3] Veracode et al. v. Appthority, 12-10487-DPW