Software written by software – for real this time
April 16, 2010
In earlier posts on this blog, I have written about who owns software written by software. It is my position that such “Written Code” falls outside the protection of copyright.
This is because the Written Code is generated by Writing Software, without human intervention on the actual choice of code used to express the functionality.
Human intervention is limited to instructing the Writing Software on the details of the functionality (the Specs).
Under copyright, only the actual Written Code could be protectable, not the functionality described in the Specs.
And since the actual Written Code is not written by a human being, and there is no human creativity involved in generating the Written Code, there is no copyright. The fact that there may be creativity involved in providing the instructions to the Writing Software, is irrelevant in respect of the Written Code.
It’s applying the distinction between an idea and its expression.
This week, I was asked by a client to review the licensing model of a software product.
It turned out that the client/provider was also offering a license to a development tool to its end-users. This tool enables the end-user to create additional functionality to an existing product (which must already be licensed from the client/provider).
An end-user challenged the IP ownership of such additional functionality to be created by the tool, and wanted to own the IP (rather than the client/provider, who standard claims IP on any functionality created by the tool, even when such functionality is created by end-users using the tool licensed by them from the client/provider).
However, according to my theory on who owns software written by software, the question is moot from a copyright point of view. Nobody owns copyright in the functionality created by the tool , because there is no copyright. Such functionality is expressed through Written Code, i.e. created by artificial intelligence.
Besides, it would be unreasonable for the end-user to ask for a monopoly on additional functionality, added to a standard product. Copyright does not cover functionality in software, it can cover only the code through which functionality is expressed (if such code is written by a human being).
By analogy, you cannot claim a monopoly (IP rights) on a specific macro you’ve written in Excel – anybody can recreate such macro.
So far, so good (although the end-user may turn out to be unhappy about this).
But what happens when the client/provider starts to use this tool exclusively in the development of a whole new application? Does that mean that such application cannot be licensed under copyright?
Well, yes.
I wonder how many software companies out there are licensing stuff that has no copyright protection?
Software written by software – help, no copyright!
April 12, 2010
No copyright!
As we’ve seen in my last blog, my position is that software written by other software falls outside the scope of protection of copyright.
Why?
Because no human intervention occurs in the actual generation of the code itself, which means no human creativity intervenes at that level.
There is human creativity in the instructions given to the Writing Software, but those instructions are ideas or algorithms, which are not protectable under copyright.
What does that mean?
It means that any code (or content, for that matter) originated by software, falls necessarily in the public domain in terms of copyright.
Does that mean that nobody owns it? Well, not exactly.
It would still be possible for the licensor of the original Writing Software to determine in the license to the Writing Software who would be owner of the resulting code. However, that is a very awkward position. It would be like Apple (can’t always use Microsoft as the baddies) saying: any automated part of your song generated through use of Garageband is owned by Apple.
It wouldn’t be fair, and it probably wouldn’t work.
It wouldn’t be fair, because Apple is not actually involved in the production of the music. And of course, what would be next? Microsoft owning the content of this blog, because I wrote the first draft on Word software (sorry, force of habit)?
It wouldn’t work, because since the music created by Garageband would fall outside copyright (not created by a human), Apple would have no efficient legal tools to enforce their position. The reason for that is that the legal tools given to copyright holders are much stronger than the legal tools available under contract – in short, it would most probably not be cost efficient to start suing people because they are in breach of a contract stating who owns code or content.
Does that mean that software becomes unprotectable?
Some of it, yes.
Is that a problem?
Not necessarily. The best explanation is – again – in the EU legislation on applying copyright to software. The first relevant consideration states literally:
“Whereas (…) the development of computer programs requires the investment of considerable human, technical and financial resources while computer programs can be copied at a fraction of the cost needed to develop them independently.”
Let’s think for ourselves, rather than accepting a piece of legislation as truth from above.
Is this consideration, which is absolutely essential to the underlying reasoning of applying copyright to software, still true?
Well, when software develops other software, the marginal cost of developing another application quickly drops to zero as well!
It’s like a derivative of Moore’s Law: the cost of an application or functionality drops by factor x over y time (someone ought to check it).
If developing an application costs close to nothing, why apply a legal monopoly with a duration of at least 70 years to protect anyone from copying it, as copyright does? Does legislation like this, based on a fundamentally incorrect premise, still make sense?
The best illustration that there is a whole field of software being developed that has little to no value in the market, and that typically represents basic functionality, is the enormous market of applications that has sprung to life for smartphones (Mac-, Android or Symbian based).
Most of the “simple” applications are offered at no fee, or a very small one.
Again, what’s the point of throwing the mantle of copyright, including its criminal liability and heavy penalties, on something that, although easy to copy, has very low development cost, and is generated by software?
So – a bit of a paradigm shift, and we’ve only seen the beginning, as this article in the Guardian shows
Does that mean that all software becomes unprotectable? No. As with content (and technology), high-end and high-value functionality will still be protected, and charged for. It’s simple economics really: if it has high enough value, it should be protected, and the law should provide the right tools for that protection. However, the borders will continue to shift.
Next time, we’ll look at two more interesting questions raised by all this:
- In the heel of the hunt, who is the author/owner of software or content created by software?
- What will be the position on anything created by artificial intelligence (AI)?
Joren De Wachter
