As more and more companies join the race to deliver the best product or service on the market, we ask ourselves how the use of AI is likely to impact intellectual property ownership.
To date, the intellectual property rights of an author in designs or products can be protected through Copyright and Patents. Copyright offers protection of the expression of something; for example, a story, music or piece of art. It’s automatically owned by the author of the piece of work in question. But who is the author?
In the past, this question has been a straightforward one, with the Copyright Designs and Patents Act 1988 defining an author as the person who ‘creates’ the piece of work. However, how does this work when Artificial Intelligence is used to create something?
Recently, the first–ever movie trailer created using Artificial Intelligence was launched to promote the movie ‘Morgan’, an AI–based thriller due to hit the cinemas this month.
20th Century Fox approached IBM to ask if Watson could help them create a trailer. Watson duly pulled the most highly emotive scenes of the movie together to create a hard–hitting trailer. A human employed by IBM then went through and pieced together the clips that Watson had selected, adding the essential ‘artistic’ element. If you haven’t seen the trailer, you can see for yourself the outcome of an AI-produced movie trailer here.
So, who (or what) created the trailer? Maybe it seems clear that the individual created the trailer, but they only had the clips because Watson had picked them out of the movie. However, Watson was only able to extract the most stirring moments because someone had fed Watson a large number of highly emotive clips for it to learn what the key emotions looked like.
So, who really did the legwork in the creation of the trailer? Was it the individual who pieced it together or the person who fed Watson the data? Could ownership actually belong to the people displaying those key emotions in the photos that Watson was fed – after all, without them Watson wouldn’t have been aware of what these emotions looked like? Or could IBM, the creators of Watson, stake a claim?
The CDPA 1988 states that, where work is ‘generated by a computer in circumstances such that there is no human author of the work’, the work will be deemed to have been created by the person ‘by whom the arrangements necessary for the creation of the work are undertaken’. This doesn’t leave any room for the AI platform itself to be considered as having ownership rights.
However, in a situation where AI is involved, it could be difficult to identify a single creator.
The situation is equally unclear when it comes to patent protection. Patents can be granted to protect designs and plans; the way things work. Unlike copyright, filing for a patent is a lengthy process. However, there are rules in place in the event of a dispute over patent ownership, but these rules don’t reflect today’s digital society and the possibility of the use of AI that is it brings.
Given that the CDPA was written in 1988 – before the iPhone was invented and before the World Wide Web existed – in a time where Artificial Intelligence was purely the domain of seemingly far-fetched Sci–Fi movies, we have to question its relevance when it comes to determining the ownership of creations that have relied upon the use of AI.
The issue of ownership isn’t the only area which will remain unclear until the law is updated. There are many ways that the law, as it stands, will be tested as AI opens up new eventualities such as the use of robots in care environments, driverless cars and patient diagnoses and operative support. As Katie Gibbs, Head of AI for IT and Business consultancy BJSS, concludes in her article ‘What does AI mean for the legal profession?’, the advent of Artificial Intelligence will demand a whole raft of legislative changes to ensure rights are protected in this modern world.
If you would like to discuss this further please contact me firstname.lastname@example.org or on +44 (0)118 907 3075.Back to Articles