Copyright, The Internet & Generative AI – “It Was On The Internet, So It Doesn’t Belong To Anyone…Right?!”

I expect that you’ll have heard someone say this before. You might even have said it yourself.

After all, the internet is a vast and publicly available resource from which many of us get our news and entertainment. It’s how we listen to music, shop, and post updates about our lives.

With that in mind, it might seem logical to conclude that what is available at the click of a button is fair game, for you to use and copy, if you so choose.

There’s nothing physical to take – it’s all just out there in the ether – you’re not doing anything wrong in downloading that film to watch at home, right?

Wrong: unless you have the requisite permission to download that film, (for example, through your Amazon account) then, more likely than not, you’ll be infringing someone’s copyright.

A copyright infringement can occur in several different ways, but the common thread (as the name suggests) is that unauthorised copying must have occurred. Downloading a film to watch creates a copy of the film – this is unlawful, unless you have permission from the copyright owner, to do so.

Unlike some other forms of intellectual property, there is no need to register the copyright for it to exist – the requirements for protection are either met, or they are not.

Firstly, the work must be a ‘work’ recognised by the laws of England and Wales. Currently, this is a finite list – of literary, dramatic, musical, or artistic works, sound recordings, films and broadcasts, and typographical arrangements of published editions. (Watch this space as, due to a change in case law at EU level pre-Brexit, it is certainly possible that a more flexible approach will be taken by the English Courts).

Secondly, the work must be original to the author of the work. For copyright purposes, this means that the work must be the ‘author’s own intellectual creation,’ in that it reflects the author’s personality. Works dictated by technical considerations or other constraining factors will not qualify because, ultimately, copyright is about creative expression.

Thirdly, the work must have been ‘fixed’: this means that the expression must be in a permanent form. You cannot protect an idea for a film through copyright, but the screenplay that you type out will qualify for protection.

Finally, requirements for qualification must be met. A work can qualify through its author’s (the person who created the work) nationality or residence, the country in which the work was first published, and (in the case of a broadcast) the country in which the broadcast was made. In each case, qualification is with reference to Britain, or another country on whom reciprocal protection has been conferred by a statutory instrument.

AI-generated content

But what about AI-generated works. Does copyright subsist in those works and if not, should it?

In the UK, copyright law does at first glance make provision for this situation in some circumstances: the author of literary, dramatic, or artistic works that are computer-generated is taken to be the person who made the necessary arrangements for the creation of the work (section 9(3) of the Copyright, Designs and Patents Act 1988).

So sorted, then?

Not really: the 1988 Act was very much of its time, reflecting the technology as it then was.

Consider the wording of the statute a little further:

Put another way, considering the fundamentals of copyright law referred to above – whose intellectual creation and personality is reflected in the output? Is the AI algorithm a constraining factor that negates originality?

This issue is yet to be litigated in the UK but no doubt it is coming.

Until then, maybe hope that whatever it is that you’ve taken from the internet is AI-generated and keep your fingers crossed that you’re not a test case.