Who owns software written by software (part 2)
February 9, 2010
In my last post, I raised the question “who owns software written by software?” based on the observation that there is a new development in the software market: applications that are fully written by other software (“Writing Software”), and that it is not quite clear who would own the result (“Written Code”).
Why is it important to know who owns Written Code? Because most software business models are based, partly or completely, on the premise that a licensee needs permission, and therefore, must pay for the right to use the application.
This business model can only work if it is clear who owns the application licensed.
In order to answer our question “who owns software written by software”, we need to understand the status of Written Code. In order to do so, let’s look at which Intellectual Property Rights (IPRs) can apply to software in general.
Typically, software is protected through three possible IPRs: copyright, patents, and contract.
Copyright is by far the most important. Why? For three reasons.
First, the general rule is that any code (written by humans) is protected by copyright. This protection is automatic, and copyright vests in code written by humans without any need for registration or other instrument. Therefore, pretty much all software code is protected by copyright, without any costly registration or acceptance process.
Second, copyright gives much wider protection rights for the owner of the software to act against unauthorized use. If a user breaches copyright (rather than a simple contract to use the software), the owner/licensor has much more efficient legal actions available under the different legislations (e.g. in the US the Digital Millenium Copyright Act, in the EU, the different transpositions of the Directive on the enforcement of Intellectual Property Rights). The owner can ask an injunction, or seize counterfeit, etc…
(Incidentally, this is also the technical legal reason why open source has been successful in the courts: copyleft is really a very specific and clever use of copyright, rather than its opposite.)
Third, most software licenses are based on copyright only. This is why software licensors get to apply all those restrictions, like user-based or CPU-based pricing, which are typically essential parts of their business model.
What about patents?
Actually, patents are used a lot less to protect software, and this is perfectly logical, for three main reasons.
The first is that the legal status of patents on software is much less secure than copyright. This is illustrated by the endless discussions in Europe, where the European Parliament killed the proposal to set up an EU Patent System (to replace the current EPO system, which is not EU, but broader European, and not very efficient) because of the hefty discussions around software patents, and the uncertain outcome of ligitation in the US, where the Supreme Court will have to decide on the validity on patents on business methods through software (re Bilski).
The result, from a business perspective, is that when you file a patent on software, you’re not actually certain if it’s ever going to stand up in court.
Second, patents are much more expensive. You need to file, register, pay every year, in a great number of jurisdictions. (Copyright, on the other hand, is obtained without any cost).
Third, patents give a very narrow scope of protection. I’m shortcutting a whole technical discussion, but if someone else develops a similar functionality as your patented one, they can avoid infringement by making the right kind of small changes to it, still operate in your market and target your customers.
In general patents on software are much less cost-efficient than copyright.
Third, it is possible to protect software through contract (that’s not technically an IPR, but contracts can give protection). While it is perfectly possible to state in a contract “I own this software, and this is what you can do with it”, the reality is that if the contract is not supported by underlying copyright or patent in the software, the legal protection available to the licensor is in practice very weak.
In order for a licensor to rely on contracts to enforce protection, there are a great number of (costly) hurdles to be taken: you have to provide evidence of breach, you need to establish damage, and it’s very difficult to get an injunction or other short-cut legal enforcement tool. In reality, mere contractual protection would only work for very valuable software, in B2B relations between larger players. It is totally cost-inefficient in any market with larger numbers of software being distributed.
So, we’ve established that copyright is the most important IPR to look at, in order to understand the status of software written by software.
The obvious next question is: is software written by software protected by copyright?
I’ll address that question in a next blog, which will centre around whether Written Code is creative work.
Watch this space.
Joren De Wachter
