Nvidia, a market leader in the gaming and technology industry for its revolutionary “GeForce” GPUs (widely used in gaming PCs and laptops), DLSS “upscaling” technologies and AI machine learning hardware, has been unsuccessful in its attempts to register UK patents for its Neural Gaming Technology and AI Medical Technology. In this article, we consider the issues that Nvidia faced, and what this means going forwards.
When one hears the word “gaming,” one often associates with major console manufacturers such as Sony (PlayStation), Microsoft (Xbox), and Nintendo (Game Boy, Switch). However, behind these consoles are technology companies whose advanced innovations power these systems and enable them to deliver cutting-edge performance. Among the most prominent of these companies is Nvidia, whose graphics and computing technologies have played a pivotal role in shaping the modern gaming experience.
Nvidia’s grasp on the gaming market cannot be overstated; its GeForce range of graphics cards are considered the pinnacle of gaming technology, and its latest RTX 5090 graphics card retails at £1,889.00 (as at the date of writing). Furthermore, its DLSS upscaling technology (which uses AI technology to improve frames rates and image qualities in games) is considered industry-changing and is now starting to be followed by other gaming/technology companies.
It therefore comes as some surprise that its recent technologies were not permitted to be patented by the UK Intellectual Property Office.
Neural Gaming Technology – 01 July 2025
The use of AI in modern society is now prevalent, and Nvidia is no exception. One of its AI technologies is the use of neural networks to generate personalised video game recommendations for players; primarily to improve their gaming skills. To do this, the technology considers:
Predicting which games a player may enjoy.
This, in turn, enhances user/player engagement and assists them in discovering games they are likely to enjoy.
Even with its unique (and innovative) technology, Nvidia’s attempts to patent this at the UK IPO was unsuccessful. The UK IPO’s decision (here) cited the following reasons for refusal:
Further, in wishing to obtain a narrower version of its patent (being the recommendations of gamers to take breaks based on their emotional condition or alertness), the UK IPO stated:
“The invention… simply provides a recommendation to a player of a game to take a break based on their emotional condition or alertness. Any effect is merely on the player being able to better progress through the game, and for the reasons detailed above… this is not considered to provide the required technical effect.”
In effect, although the AI technology was considered inventive, it was deemed to relate to "excluded matter" due to its lack of technical contribution beyond providing cognitive content. This deficiency arises from the requirement that an invention must offer a technical solution to a technical problem, rather than merely constitute a computer program. The UK IPO concluded that the technology in question did not meet this threshold, as it merely provided recommendations or advice to players without improving the underlying technology or the operation of the computer itself.
AI Medical Technology – 27 June 2025
In a different, though related, UK patent registration, Nvidia developed a neural network that utilises medical imaging and clinical data to enable the efficient allocation of limited medical resources. For example, CT scans and clinical data could be used to train the neural network to predict whether patients with COVID-19 would require oxygen therapy, including estimating the need for an intensive care unit bed.
The effect of this approach is clear: to ensure that limited medical resources are allocated appropriately, providing the necessary supplies to patients at the right time.
However, similar to Nvidia’s neural gaming technology, the UK IPO rejected Nvidia’s claim, stating that the AI Medical Technology merely provides an advisory recommendation without a technical effect. It concluded that the invention does not solve a technical problem or contribute a technical innovation, as it only computes likely treatment effectiveness and supports administrative decisions rather than delivering a technical solution; and that:
“Had embodiments instead referred to the output as specifying an exact amount of a medicine to prescribe or settings to apply to medical apparatus to apply treatment my conclusion may have been different.”
Despite the above hypothetical, the invention nonetheless falls within the excluded categories due to its lack of technical contribution, as required under UK patent law.
There is already extensive commentary on the interplay of AI and intellectual property rights. What distinguishes Nvidia’s case, however, is the need to look beyond mere AI application and question whether the use of AI to train proprietary systems constitutes a registrable intellectual property asset.
The UK IPO has clarified that for AI-driven innovations to be patentable, they must demonstrate specific, tailored solutions that produce a clear “technical effect.” Yet, given the ongoing evaluation of AI’s broader societal impact, it remains challenging to predict when, and if, the UK IPO will fully embrace AI as an integral component of registrable technological innovation.
Culbert Ellis specialises in Intellectual Property law. If you would like to discuss how we might help protect your brand, please contact Wing Ming Choi at wingming.choi@culbertellis.com or call +44 (0)203 987 0222.
