AI and generative product design: An increasing risk for IP?

What is ‘Generative Design’?

Generative design is a cutting-edge approach in product design which uses data, algorithms, cloud computing and AI to autonomously generate optimised design solutions based on predefined goals and constraints (such as the types of materials, manufacturing methods, weight limits, costs, strength requirements). When compared to a ‘traditional’ design method - which typically derive from a manufacturers experience, knowledge and creative ideas – generative designs offer greater efficiency (time and costs), sustainability, innovation and complexity.

Essentially, this fundamentally means that manufacturers are no longer required to create and design their own products manually and generative design operates as collaboration process between manufacturers, designers (who define the parameters and constraints) and software (which generates the designs). 

Intellectual Property Rights

When considering manufacturing design, there are a variety of Intellectual Property (‘IP’) rights that could arise, including but not limited to copyright,  design right and database right, but also quasi IP rights relating to confidential information and trade secrets. Finally, it is also possible, at least in theory and dependent on its subject matter, for a generative design to be an invention that is capable of patent registration.

It has recently been determined that AI cannot be the inventor for the purposes of a patent application.  Furthermore, the Patents Act 1977 and its European counterpart, the European Patent Convention, restrict the patentability of computer programmes, so there is a fine line between what is and is not patentable.

What are the risks to manufacturers?

Despite the endless possibilities that generative design offers manufacturers, the process is not without risk.  

Due to the algorithmic nature and reliance on generative design, there is a possibility that a generated design may be similar - or identical to - designs that are already in existence.  There are also copying considerations at play, given that in order to operate effectively an AI model often needs “feeding” with information.   Not only does this pose risks of infringement, but it can also impede a manufacturer’s ability to demonstrate the design is novel and involves an inventive step (which could be detrimental to its registration, whether as a patent or design).  

Additionally, given the generative nature of such designs, it often means that they are highly intricate and complex designs. Another risk to consider is ensuring that any graphical representation fully captures the complexity and nature of such design and its features; given the particular importance of accurate representation of a generated product when presenting a suitable two-dimensional or three-dimensional design for protection.

How can we assist?

At Forbes, our specialist IP team has extensive experience and expertise in assisting manufacturing businesses with a variety of complex needs when identifying their designs and protecting them against third-party copying or use without consent. We also assist in drafting and negotiating research and development agreements, as well licence agreements once created.