AI in manufacturing and engineering and intellectual property considerations

At its' core, artificial intelligence (AI) is a form of computer science which involves the development of programmes to enable tasks to be completed which would otherwise require human intelligence. AI algorithms can be used for tasks including problem solving, logical reasoning and understanding language and are often trained to respond to a particular input in a certain way. Many may be aware of the presence of AI, for example, in online shopping recommendations or GPS navigation systems, however it is also widely used in the manufacturing and engineering sectors. Some examples of how AI is used in such sectors include:

• Detecting defects on the production line, reducing or in some instances eliminating the need for hours to be spent on quality control by employees;

• Assuring the required quality of goods via image processing algorithms which automatically validate whether an item has been produced to an adequate standard. Such technology often have cameras which can detect even the smallest of defects which may not be noticeable to the naked eye;

• Forecasting the price of products and materials;

• Generative design, for example where a designer inputs information such as design criteria, overall goal, and cost constraints, into algorithms which explore the possible permutations and generate design options via machine learning techniques;

• Utilising robotic machines for predicting when machines need maintenance and upgrading, thus preventing unplanned downtime.

AI is becoming further accessible for more businesses and though it is not anticipated that AI will wholly take over the need for humans, it can be useful in many instances to capitalise on the advances which have been made, to not only reduce costs but to also enhance and speed up processes.

An interesting consideration for those utilising AI is the intellectual property rights subsisting in the AI itself, but also in any inventions subsequently created by AI.

Patents protect inventions, such as new processes which AI may create. Presently, only a human can be identified as an inventor in a patent application and this was reflected in 2019 in a series of patent applications made to the Intellectual Property Offices of the UK, EU and US by Stephen Thaler who claimed that an AI system, DABUS, had produced an invention and should be listed as the inventor. All three Offices refused Thaler's application on the basis that patent inventors must be human.

To contrast, under the Copyrights, Designs and Patents Act 1988 (CDPA), copyright protects various categories of works, including literary, dramatic, musical or artistic works, including software code, computer programmes and databases which can be created by AI. Whilst traditionally software is developed by a human, meaning that the owner of the copyright in the software would be the developer, questions arise with respect to copyright ownership of works created autonomously by an AI system, and much will depend on the extent of human input involved.

The CDPA accommodates 'computer-generated works', and as such if there is no direct human author, the author is the person who undertook the arrangements necessary for the creation of the work. In straightforward cases, the organisation controlling the use of an AI system is likely to own the output, however there are often multiple parties involved with AI systems thus making it unclear who has made the arrangements necessary for the creation of the work.

The position under intellectual property ownership and originality in respect of AI is therefore not always clear-cut and may give rise to disputes. In the absence of certainty under legislation, it is important to ensure that AI-based projects are governed by contracts, clearly defining the ownership of intellectual property rights subsisting at the time the contract is entered into and beyond.

For assistance with dealing with the intellectual property rights arising from AI systems or ensuring your business has in place robust agreements to deal with such systems in general, contact John Pickervance, Partner and Head of Commercial.