Copyright – Protection for Software

How can software be protected?

There are various ways the intellectual property of software that has been developed by a business or person can be protected.

The high level of the criteria required to patent software is difficult to meet – a much easier way to obtain protection lies with copyright. Copyright protects your work and stops others from using it without your permission. It arises automatically and it’s free.

Copyright protection on software stems from the Copyright, Designs and Patents Act 1988 (CPDA 1988) and Directive 91/250/EC (the Software Directive).

What sort of protection does copyright afford?

CPDA 1988 specifically provides copyright protection for computer programs, preparatory design material for a computer program and databases. To attract copyright protection, the work must originate with the author and there must have been ‘a modest amount of skill, labour or judgment’ involved in its creation.

Copyright does not subsist in a work unless and until it is recorded, in writing or otherwise. Any ideas or principles underlying the subject are not therefore protected and should be protected by patents if applicable.

Copyright prevents others from:

  • copying your work;
  • distributing copies of it, whether free of charge or for sale;
  • renting or lending copies of your work;
  • performing, showing or playing your work in public;
  • making an adaptation of your work;
  • putting it on the internet.

The closer to the original the copy is, the stronger the infringement. This means if the copy has a substantial amount of the original work directly copied into it, the infringement should be relatively easy to prove.

Under s 21 of CDPA 1988, the making of an adaptation of the work is an act restricted by the copyright in a literary, dramatic or musical work. In relation to a computer program or database, this means an arrangement or altered version of the program or a translation of it. The problem, when it comes to copyright in software, is non-literal copying.

Non-literal copying

Non-literal copying is a form of copying that does not include the wholesale lifting of one program into another. It can arise where elements of the program such as its arrangement, chain of operations, functions, interfaces and styles are copied but the program code is not directly copied.

One problem lies with assessing how much the substantial part of the software has been taken, bearing in mind that the law gives protection for the expression of a work not the idea behind it. In term of software, this idea/expression dichotomy means the code that is written is protected but what the programme does or how it achieves its goals is not.

How do the courts approach this?

Access plus similarity

The earliest cases in this area took place in the US, with Europe using these cases to inspire their own protection measures. In the US case of Whelan v Jaslow a similar program was written in a different programming language. There were step-by-step similarities in essential subroutines, leading to the suspicion of copying. In this case a very simple test was used. It first had to be shown the defendant had access to the software in question and then that there was an ‘objective similarity’ between the defendants program and the original.

Access may be found if, for example, a programmer has switched companies or is a freelance contractor who has worked for both the copyright holder and the defendant. If this form of access has been found it may be possible to further sue the contractor or ex-employee for breach of confidence.

Abstraction-filtration-comparison (AFC)

The test firstly reverse engineers the programme, identifying the key elements (an abstraction), secondly it removes the key elements that are common in the art (filtration) and finally compares what is left of the two programs to see if there is a substantial taking (comparison). This was used in the UK in the case John Richardson v Flanders.

Over-borrowing

This test was introduced when it was thought the AFC test was too complicated. This test is similar to the idea of looking for access plus similarity but the elements of the program including the code, the structure and the design are also taken into account when comparing the two programs. For example, common spelling errors throughout the programming, common comment headings and similar redundant sub-routines would be looked for. This approach has been used in cases such as Ibcos v Barclay and Cantor Fitzgerald v Tradition UK.

What is the current situation?

Non-literal copying is still a contentious field. In the UK it appears the test of access plus similarity still holds a lot of weight. It also seems the AFC test preferred in the US is, in the opinion of the UK courts, too complicated. A simple test of over-borrowing was enough in Ibcos v Barclays. It must be remembered, however, that any infringement will still be assessed under the traditional substantial parts test from conventional copyright, after the work to be protected had been found.