Czech court finds that AI tool DALL-E cannot be the author of a copyright work

In a recent judgment (in Czech) which is one of the first of its kind in Europe, the Municipal Court of Prague (the Court) held that an image generated by an AI tool was not capable of being protected by copyright, as it was not authored by a natural person.   

Background

The claimant in this case, which is anonymised in the Court's judgment, had asked the AI program DALL-E to create an image for the claimant's website. DALL-E is a text-to-image model developed by OpenAI to generate digital images from natural language descriptions, called "prompts."

The prompt used by the claimant in this case had been: “Create a visual representation of two parties signing a business contract in a formal setting, such as a conference room or a law firm office in Prague. Show only hands.” The claimant subsequently placed the image which had been generated by DALL-E (the Image) on its website. 

The defendant, a local law firm, subsequently copied the Image and used it on its own website (presumably, to prove a point?). 

The claimant brought this action for copyright infringement against the defendant, alleging that it was the author of the Image, and seeking injunctive relief. 

The Court’s judgment

The main issue which the Court had to determine was the authorship of the Image, and whether it was possible for AI to be the author of a copyright work. 

The Court noted that the Czech Copyright Act (the Act), and particularly Article 40 thereof, confers upon the author of a work the right to demand, amongst other things, a determination of their authorship of the work and various forms of injunctive relief aimed at prohibiting unauthorised exploitation of the work. 

Article 5(1) of the Act states, however, that the author of a work for the purposes of the Act is "the natural person who created the work".

In this case, it was not disputed by the parties that the Image was created through the use of DALL-E, but the claimant's case was that DALL-E created the Image based on their specific instruction, and that they were therefore the author of the Image. 

However, the Court noted that the claimant failed to substantiate this allegation within any evidence (beyond their own witness evidence), and consequently failed to meet the burden of proof for showing authorship. The claimant did not therefore have legal standing to bring the claim in question. 

The Court also observed that the Image (being an image created by AI) could not constitute a work capable of copyright protection as it was not the result of the creative activity of a natural person pursuant to Article 2 of the Act. The Image was not therefore a copyright work, let alone a copyright work owned or authored by the claimant. 

Unaltered photo of Kats conducting a business meeting

Comment

Whilst the decision is not particularly surprising, and is broadly consistent with the handful of other decisions we have seen on this topic in the past few months, it is interesting in that the Court did not altogether rule out the possibility that the plaintiff could have authored the Image, provided they had adduced sufficient evidence to show this.  

This raises questions about what such evidence might consist of (noting that the claimant's own witness evidence was insufficient) and what extent of human direction/involvement needs to be evidenced in order to displace the relevant burden of proof. 

In this case, the prompt used and the method of creation was basic enough to make the resulting finding uncontroversial, but it is to be expected that the courts will eventually have to decide when the line is crossed for a work to have been authored by an individual with the mere assistance of AI (which, this Kat expects, would be protectable, assuming sufficient creative control/direction is retained by the human author).  

In the meantime this decision serves as a reminder that, particularly in Europe, anyone using images generated by AI tools such as ChatGPT and DALL-E can expect not to be able to rely on copyright law to prevent the copying of such images by third parties, and instead to seek to put in place other (e.g. contractual) protections to prevent this.