Specialist advice on patenting AI inventions in the UK.
Artificial intelligence is creating commercially valuable technology across software, search, recommendation, data processing and technical systems. Patentability, however, still depends on the invention itself, the way it is described, and how the claims are drafted. We advise businesses on AI patentability in the UK, including inventions involving machine learning, artificial neural networks, software enabled products and other computer implemented technologies.
We are particularly well placed in this area because of our involvement in Emotional Perception AI v Comptroller General [2026] UKSC 3, the landmark Supreme Court decision on AI and computer implemented inventions.
Why AI patentability is not straightforward
AI inventions often sit at the boundary between software, data handling and technical functionality. The key issue is therefore not simply whether a system uses AI, but whether the claimed invention, viewed as a whole, makes a relevant technical contribution, and whether the features relied on for patentability support that analysis.
That distinction matters because not all AI related subject matter is treated the same way. The decision altered the approach to exclusion for computer implemented inventions while keeping the focus on technical character, feature interaction and inventive step.
What changed after Emotional Perception
The Supreme Court held that the Aerotel test should no longer be followed, accepted the any hardware approach for assessing exclusion, and emphasised that once the exclusion hurdle is cleared there must be a careful assessment of which features contribute to the technical character of the invention as a whole.
That is particularly important for AI applicants because many inventions involve a mixture of technical and non technical features. Careful framing of the invention and precise claim drafting remain central to the outcome.
AI inventions that may merit advice
- Artificial neural networks and trained model implementations
- Machine learning systems used in search, recommendation or prediction
- Embedding based or vector based retrieval technologies
- AI enabled data processing, classification or signal analysis systems
- Hybrid products combining learned models with hardware or other technical means
How we help
Advice is tailored to the stage of the technology and the commercial objective. That may be an initial patentability review, drafting strategy for a first filing, advice on prosecution and objections, or a broader portfolio review where AI functionality is central to the business.
It is particularly valuable where the invention sits close to historically difficult areas such as neural networks, software led architectures or data centric systems, all of which reward careful drafting and an objective analysis of the invention as a whole.
For more information, please contact Bruce C. Dearling, Shareholder at Hepworth Browne, at bcd@hepworthbrowne.com.
