The Supreme Court considers whether an Artificial Neural Network can be patentable

Estimated reading time: 3 minutes

Can a type of artificial intelligence (AI) known as an artificial neural network (ANN) be patented under UK law?

An ANN is an AI machine learning model that takes inspiration from the way the human brain is structured by using artificial layered and connected nodes as its ‘neurons’. 

The Court of Appeal previously found that Emotional Perception’s ANN for recommending media files (e.g. a song, video or text) by modelling how a user may emotionally respond was not patentable due to falling within the exception for being a computer program “as such” (see our previous article here).

Recently, it was the turn of the Supreme Court to consider this case.

Supreme Court Decision

The Supreme Court agreed that the ANN was a computer program.

It held that a computer program included “a set of instructions capable of being followed by a computer…to produce desired manipulations of data” but such results were not necessarily predictable as computers could produce calculations and results that humans could not. Therefore, regardless of whether the ANN was executed through software, embodied in hardware or derived its functioning from training rather than specific programming, it was essentially “a set of instructions to manipulate data in a particular way so as to produce a desired result” and was a computer program.

The next step was to consider whether it was a computer program “as such” to be excluded from patentability.

The Court considered the case law and clarified that the correct approach includes assessing whether the claimed invention has technical features beyond the computer program itself.

The Court noted that the claimed ANN system involved computer hardware on which the ANN operated as well as a database for storing data files, a communications network and a user device. Therefore, it considered that the claim included technical features beyond the computer program.

As the reformulated approach had not been applied before, the Court sent the matter back to the original Hearing Officer of the UKIPO to decide whether the invention could meet the remaining requirements for patentability under the reformulated approach.

Why this matters

The decision has practical implications for businesses developing or protecting AI-driven technology:

  • Software and AI related inventions (including those based on ANNs) potentially can be patented – However, applications should be drafted carefully (and existing filings reviewed) considering the exclusion for computer programs “as such”.
  • Further guidance is possible – The case has been sent back to the UKIPO, and additional guidance may be provided once the Supreme Court’s reformulated test has been applied (and potentially tested further on any appeal).
  • AI related litigation is increasing – The Supreme Court previously decided that an AI could not be named as the inventor of a patent (see our article on the Thaler case here) and the High Court’s first generative AI decision resulted in a limited trade mark infringement finding for Getty Images (see our article here), which is reportedly being appealed.
  • Ongoing regulatory developments – The UK government was aiming to publish an economic impact assessment and report (see here) on the options in its copyright and AI consultation by 18 March 2026 (see our article providing background on the consultation here), although reportedly this may be delayed.

Organisations investing in AI-driven technology should reassess both pending and future filings. Our intellectual property team works with technology businesses at all stages of development to structure strong patent strategies aligned with commercial objectives.

Portfolio Close
Portfolio list
NameCVEmail
No one added to your portfolio yet.

Remove All

Download

Close

Exclusive Content

Unlock 24-hour access to our exclusive content including on-demand webinars by submitting your details using the form below:

"*" indicates required fields

Name*
Agreement*