Teal blue graphic

Who owns the IP rights arising from Artificial Intelligence?

Recent developments in Artificial Intelligence have raised a lot of legal questions over Intellectual Property rights. We consider who owns the copyright.

Artificial Intelligence (AI) and robotics are not new, but with recent developments in technology and the rise in computer processing power we are seeing greater advancements in this area.

The Future of Jobs Report (following last year's World Economic Forum) estimated that AI and robots could replace 5.1 million jobs by 2020.  We have already witnessed the first robot being appointed as a creative director (for McCann Erickson) and, with the continued trend towards a commitment to developing AI, the advancements in this area will undoubtedly continue to change the way in which we live and work.  

If they have not already done so, organisations looking to capitalise on the potential offered by AI should also start to think about the intellectual property rights arising in relation to it.

Considering who owns copyright

The default position on copyright in the UK is that where a work attracting copyright is created by an employee in the course of his or her employment then the copyright in such work will automatically be owned by the employer.  Ownership of work created by non-employees (such as contractors) should be dealt with by way of agreement. If organisations choose to adopt robots rather than employ or otherwise engage humans to carry out tasks, then we should consider the implications of this from an intellectual property law perspective and ascertain whether all works created by robots will gain the benefit of copyright protection.

In the UK, if literary, musical or artistic works are computer-generated the author is taken to be the person that makes the arrangements necessary for the creation of the work undertaken. The position therefore seems straightforward therefore when dealing with works created by humans with assistance of AI or robots. But what is the position if works are created by robots independently of humans?

Who owns machine created content?

'LnH: The Band' is a robot powered band that uses AI tools and Twitter to automatically generate music in real-time based on predetermined models for different music genres.  It works by combining the intelligence of maths, the laws of musical theory, training data, creative experiences and a number of user specific behaviours.   In trying to work out what IP is generated and by whom using LnH: The Band, the question we need to consider is whether copyright arises in the music and, if it does, who owns the copyright in the work generated?  Who undertakes the necessary arrangements for creating the music, a human or a machine?

In this instance, the input of the tool is controlled by the development team, who have crafted a software programme designed to be flexible enough to allow inputted data to be manipulated by the tool without further human intervention.  Arguably therefore, although the data input side is control by humans, the data output is being 'controlled' (or at least generated) by the tool itself.  Following this, could it be that the AI itself 'owns' the IP in the output of the tool?

A recent case in the US involving a British tourist raised some interesting issues around whether work had to originate from a human in order to attract copyright. A photograph had been taken by a monkey, raising the question of whether the owner of the camera owned the copyright in the photographs or whether the works fell into the public domain. Here, the US Copyright Office decided that copyright did not arise in the photographs as the owner could not be deemed the author simply by owning the camera on which the photographs were taken.  Applying this approach in relation to LnH Band, it might be too a step to far to assume that the developer is making the arrangements for each piece of music created, which supports the notion that rights may be being generated independently of human control.

Should machines be granted legal status?

The position of authorship in relation to copyright may change if robots are granted legal status. We already know that once copyright and/or other intellectual property rights are created they can be owned by an organisation. Therefore being human already is not a prerequisite to having rights. Members of the European Parliament's committee on legal affairs will shortly decide on whether robots should have legal status as an 'electronic person' and be held responsible for their actions. If robots are given legal status, it will be interesting to see if they will be capable of owning the intellectual property rights in the work they create.  The inevitable consequence of an increasing reliance on AI will be the need for future legislation to be drafted with AI specifically in mind.

This publication is intended for general guidance and represents our understanding of the relevant law and practice as at February 2017. Specific advice should be sought for specific cases. For more information see our terms & conditions.

Insights & events View all