The reality of Artificial Intelligence (AI) raises some interesting questions: Who owns copyright in AI (robot) generated content? Could that be the robot? If so, how would they hold it? How would they enforce it?
At its heart, copyright is a monopoly right granted to the author as a ‘reward’ for their creative work. However, where that work is created by an AI robot, does the robot need a reward such that its work is protected by copyright?
AI generated content is increasingly found in social media, advertising and even mainstream news and is attractive as a concept, introducing greater efficiencies to content production. In a creative context, AI has produced artistic works which can be indistinguishable from human generated content. In a commercial context, AI is set to challenge existing business, models as discussed in our recent post Artificial Intelligence and Lawyers.
The type of AI-generated content we’re talking about here is work that is normally protected by copyright, which has traditionally been for a human (or corporate entity), as owner, to commercialise and enforce. Enforcement against infringers, who copy and duplicate work without permission, can be through civil and criminal law remedies.
What’s the current position?
The first owner of copyright in a work (an article, photograph etc.) will be the author: the person who created it (normally a human).
Where the work in question is computer-generated, existing copyright law has provided that the person/company owning or directing the use of the IT systems concerned could be taken to be the owner of the work.
However, this may not provide for, or cover, how content could be created by AI in the future. AI is likely to become more prevalent and for a party relying on AI to generate content, it’s important to have the certainty of ownership so that copyright can be enforced with confidence.
The waters can be muddied by competing claims to ownership from various parties, including developers, designers, engineers and others involved in creating the AI systems. Also, as AI becomes more complex, trying to establish the owner of content is could be increasingly difficult. Could it therefore be easier in the future to simply name the AI robot as copyright owner? Could we abandon the concept of a human, or a human directed company, as owner behind it all?
As it could become more difficult to trace the genesis of content, more sophisticated systems may have to be devised to protect AI generated content. For now however, agreeing clear contract terms defining ownership with AI system owners may be our best option. Until the machines takeover of course….
Ownership of copyright is discussed in more detail in our guidance note.