Originality in the UK means "the author's own intellectual creation" including for software generated works – the debate is over (…for now) | Fieldfisher
Skip to main content

Originality in the UK means "the author's own intellectual creation" including for software generated works – the debate is over (…for now)


United Kingdom

Back in November 2023, the Court of Appeal case of THJ Systems Ltd v Sheridan [2023] EWCA Civ 1354 settled any remaining debate over the correct test for the originality requirement for copyright to subsist in a literary, dramatic, musical or artistic work in the UK.

Since then, it has been widely publicised and was recently cited in the Court of Appeal ruling in the long-running supermarket spat between Lidl and Tesco (see our blog here - Lidl bags a bargain at the Court of Appeal). This case was in the context of a computer-generated work which will be of particular importance given the current explosion in AI. The case is therefore a bit of a "double whammy" dealing with two current hot topics.


The relevant facts of this case are as follows:

  • Mr Mitchell (the individual behind the first claimant) created software which displays financial information about the performance of options in the market and presents it in a graphic form (referred to as the "risk and price charts").
  • Mr Mitchell had been in business with Mr Sheridan (the first defendant) through an LLP and granted him a licence to use the software.
  • The first instance decision (on a matter not appealed), concluded that Mr Mitchell had successfully expelled Mr Sheridan from their LLP and validly terminated the defendants' licences to use the software no later than 25 January 2016.
  • The claimants' position was that the risk and price charts were:
    • original artistic graphic works under section 4(1)(a) CDPA; and
    • the intellectual creation of Mr Mitchell as the creator of the software and therefore the person who made the necessary arrangements for the creation of the computer-generated work required under section 9(3) CDPA.
  • The defendants' position was that the risk and price charts were the intellectual creation of the user of the software, who select options and a time period for the display.
  • The claimants alleged that the copyright in the risk and price charts had been infringed by the defendants using them in various presentations and videos after termination of the licence.


The first key matter in dispute related to the originality of the works.

Historically, the UK test for originality was that the author must have used "their own skill, judgment and individual effort". This was a low bar and any negligible amount of effort could suffice. At an EU level, there is a string of case law dating back to Infopaq which held that to be original, the work must be "the author's own intellectual creation". The EU originality test has been reiterated and refined recently, particularly in the high-profile case of Cofemel (see our various blogs discussing Cofemel - First Response: Copyright in UK designs – artistic quality no longer required, CJEU gets a handle on whether copyright subsists in Brompton bicycle, but ultimate decision hinges on Belgian court and It's a race to the finish line as Judge dismisses strike out application holding that rowing machine could be a work of artistic craftsmanship? ) - to be original, the author must have been able to express their creative abilities by making free and creative choices and the content of the work must not have been dictated by technical considerations.

There had then been some residual debate about the correct test to apply in the UK and the difference (if any) between the two tests. Most commentators and members of the judiciary in extra-judicial academic discussions have endorsed the EU test as the correct one to apply in the UK. However, there were still instances of ambiguity and judges considering both tests. In some scenarios the difference between the two tests has been seen as academic and not something which would have an appreciable effect on the outcome. Also, the reporting of this case has highlighted that some industry organisations and lay people were continuing to apply the old UK test in how they approached dealings with their own works and where they deemed copyright subsisted.

In the present case, the judge at first instance concluded that "the work of creating the look and functionality of interface including the arrangements of the tables and graphs did involve the exercise of sufficient skill and labour for the result to amount to an artistic work" (see para 214). On appeal, it was held that this was not the correct test. The judge seems to have applied the old test because he was (surprisingly) not referred to any of the relevant case law and was primarily considering two cases regarding copyright in graphic user interfaces which had been decided prior to Infopaq – i.e. Nova Productions v Mazooma Games [2006] EWHC 24 (Ch) and Navitaire Inc v Easyjet [2004] EWHC 1725 (Ch).

Giving the leading judgment, Arnold LJ confirmed that the two tests for originality were not the same (as could be seen in Football Dataco and Funke Medien) and the European test was "more demanding". He made five important points about this test for originality: (1) the test is objective, (2) it is not a matter of artistic merit, (3) the burden of proof lies on the claimant, (4) a key piece of evidence is the works themselves, and (5) the functionality of the software is irrelevant.

Points 1 to 3 may seem straightforward, but the final two points are certainly important to remember. It is easy when faced with copyright in technical, functional or informational works to become distracted by this. However, this is a helpful reminder that the key assessment relates to the visual appearance – this case was not about the level of originality in the software itself, it purely related to the graphic output.

Key aspects of the risk and price charts were that they had "been laid out with some care", "designed … so as to cram quite a large amount of information into a single screen", there were "choices as to what to put where, including such matters as which commands to put into the ribbon and in what order" and selections as to the fonts and colours. Taking these points into consideration, Arnold LJ found that the charts were original. Arnold LJ accepted that the "degree of visual creativity which went into the risk and price charts was low", but that this only meant that "the scope of protection conferred by copyright… is correspondingly narrow, so that only a close copy would infringe".

In his analysis of the originality, Arnold LJ also noted evidence which had not been adduced. This included elaborating on the choices made in the design of the charts and whether they were dictated by technical considerations or considering similar graphical user interfaces produced by third parties. It is unclear if Arnold LJ is suggesting that the defendants should have adduced such evidence, but it could be taken as a hint for future defendants to pursue these lines of attack. This approach may lead to evidence of the existing design corpus and prior art, as encountered with other IP rights such as patents or designs.  

In this case, the alleged infringement related to uses of the actual charts created through this software. Therefore, there were no questions of substantial part or an assessment of visual similarities. In cases where there are allegations of infringement which involve the creation of similar charts for example where the functionality is the same and certain elements have been recreated, the narrow scope of protection is likely to have a greater impact.


The second area of appeal related to infringement of copyright.

Although the trial judge at first instance had found that copyright subsisted, he had held that it was not infringed. This was overturned on appeal; the Court of Appeal held that the instances of uses of the risk and price charts which had been pleaded did constitute an infringement and directed an inquiry as to damages or an account of profits at the claimants' election.

The reasons for the appeal as to infringement were primarily procedural rather than technical. There were various ambiguities in the defendants' pleaded case regarding infringement. Relying on various admissions made by the defendants' legal team and the wording of the pleadings, Arnold LJ found that their only pleaded case on infringement was based on the challenges to the subsistence and authorship/ownership of the copyright i.e. the defendants were only asserting that they had not infringed copyright on the basis that no copyright subsisted and if it did, the claimants were not the rightsholders. It was therefore not open to the defendants in closing written submissions to "ambush" the claimants with a new argument that there was no communication to the public because there was no targeting in the UK. The trial judge at first instance had found there to be an "evidential vacuum" on this point and that the claimants had therefore failed to prove a claim for infringement. However, the Court of Appeal's ruling was that there was no evidence precisely because it was not a matter in dispute at trial and the judge was therefore wrong to deal with the issue of infringement.


At first instance, the defendants had denied that the risk and price charts were the intellectual creation of Mr Mitchell. Instead, they asserted that the author was the user of the software who selects the options and period of time to depict in the graph. On appeal, the dispute solely related to the originality of the works (as discussed above) and it was no longer in dispute that if they were original, Mr Mitchell was the author (and the owner was the first claimant). However, the question of authorship is still interesting to consider.

The parties had pleaded the case under section 9(3) CDPA, which states:

"In the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken."

A "computer-generated" work is defined as meaning that "the work is generated by computer in circumstances such that there is no human author of the work".

There is some debate currently surrounding the requirement for originality in the context of computer-generated works. There are some commentators who consider that section 9(3) operates as an exception to that originality requirement, arguing that the wording of section 9(3) allocates authorship to the person who finances or organised the creation of the work rather than the person who uses their own intellectual creation. Alternatively, other commentators argue that section 9(3) is simply meaningless and the normal considerations of originality still apply – and seek to rely on this case in support of that view.

Although authorship was not explicitly a matter to determine on appeal, the Court of Appeal appeared satisfied to accept Mr Mitchell as the author of the works - the analysis of the relevant aspects of intellectual creation for originality as set out above, were all the acts undertaken by Mr Mitchell in designing the user interface. The operation of the software (or even the fact that these graphics were generated using software) seems to have been irrelevant. This assessment and the references to the acts of the human author in this case may suggest that section 9(3) is meaningless given the court still applied the usual test for originality. However, another angle could be that the court did not consider the works in this case to even be computer-generated within the above definition. Afterall, the protectable elements were not those of the data inputted and the parameters set by the users and the technical output generated by the software; copyright subsisted due to the visual graphical elements in the layout which had been designed personally by Mr Mitchell.

On this basis, the present case would not provide any useful precedent for AI generated works. With AI generated works there is truly no human author as per the above definition. The output results from a combination of the development and training of the AI model (which is highly technical and unlikely to satisfy originality requirements but could still fall within section 9(3)) and the prompts inputted by the user (which may express the author's own intellectual creation, but the resulting image may not be a real reflection of that).

Therefore, the questions of authorship in relation to AI generated works is still one which will need to be addressed by the UK courts (or legislature) in the future.


Although this case does not change the law on originality, it clarifies the correct approach to be used in the UK. It has helpfully attracted a good degree of attention and interest which will hopefully assist in focussing people's minds on this higher threshold which may have a bearing on commercial dealings with copyright works. For traditional artistic works, this should not be much of an additional hurdle. However, it may be more challenging in scenarios such as here where the work is the output of software or a simple photograph reproducing a 2D work.  

Some commentators have seen the court's comment that originality does not require artistic merit to implicitly support the position that the UK's system of closed categories of copyright work is inconsistent with the pre-Brexit EU position and therefore assimilated law. Although that has been the subject of much academic discussion (including from Arnold LJ himself), it is questionable whether the brief comment in this case provides much support for that point. It could simply be considered a statement repeated over time in a number of copyright cases, particularly those regarding LDMA works. The highly anticipated judgment in the WaterRower case (see our blog referred to above) is perhaps more likely to provide interesting insight into the categories of copyright works.

This case also shows how engrained various EU doctrines are in our national laws. Even after Brexit, senior experienced judges with IP expertise are continuing to follow the established EU case law rather than reverting to the historic test. Although this case was heard before the Retained EU Law (Revocation and Reform) Act 2023 (REULA) came into force –(see our blogs for more details - Retained EU Law (Revocation and Reform) Act 2023 receives Royal Assent) ), the Court of Appeal's judgment indicates that there is currently no immediate rush to  depart from assimilated EU case law on this topic. Far from any negative treatment or implicit comments, this case is a strong endorsement that, for now, the EU approach continues to be the correct test to apply in this jurisdiction. However, it is still open for judges to take a different approach in the future using the powers under REULA…. Watch this space.

Areas of Expertise

Intellectual Property