Copyrighting Dreams and AI creations

by Kadri Bussov

In 1818 Mary Shelley had a dream that resulted in one of the most famous novels of modernity – Frankenstein: or, The Modern Prometheus.

The fascination with human cognition has spurred the development of a groundbreaking technology capable of altering the very fabric of human existence. We find ourselves on the brink of a quasi-dream state, where the notion of capturing dreams is no longer fanciful. The images showcased in my blog represent an extension of my creativity through Artificial Intelligence. It's not far-fetched to speculate that a technology capable of intercepting and recording human dreams—a 'dream-catcher'—could emerge, potentially becoming commercially available. One might even ponder whether current AI capabilities already allow for such feats. With the advent of such technologies, a crucial question arises: to what extent are our creations, extended by AI, protected?

A few years ago, I delved into this concept in a paper discussing the possibility of copyrighting dreams. The focus was on the three key prerequisites for protection within the intellectual property regime—originality, expression, and fixation.


A question of fixation

"The dream is of a first-person experience of being trapped on a radioactive planet. The only means off the planet to spaceship waiting in orbit is to catch an elusive red bird with teleportation powers. The twist that is revealed later is that the planet was the Earth all along. The dream is also being recorded for future revisions."

The Bern Convention, in Article 2(1), defines "literary and artistic work" to encompass every production in the literary, scientific, and artistic domain, regardless of its form of expression. The article provides a non-exhaustive list of various forms of expressions. Article 2(2) introduces the fixation requirement as an option for copyright protection. Fixation mandates that a work is embodied in a form that physically expresses the underlying idea.

However, the 2012 Beijing Treaty on Audiovisual Performances presents an alternative perspective on the fixation requirement for performances. According to WIPO's explanation texts, the treaty aims to enhance the protection of creators in the digital era. Some jurisdictions find creations that inherently change over time challenging to meet the fixation requirement.

In the United States, the Seventh District Court, in the case of Kelley v Chicago Park District, deliberated on the arrangement of wildflowers in a public garden. The court concluded that the arrangement of wildflowers does not meet the fixation requirement as wildflowers are inherently in constant change. This decision contrasts with the United Kingdom, where spontaneous speeches and conversations can be protected by copyright, provided they are recorded and original. Analogous arguments could apply to live streaming dreams.

The issue of authorship in the Kelley case stemmed from the fixation requirement. Similarly, questions arise concerning dreams—are they "forces of nature" or akin to spontaneous conversations? Where does artificial intelligence fit into this framework? Carpenter and Hetcher argue that fixation should only separate idea from expression, without further implications in determining copyrightable works. They contend that the fixation requirement unfairly discriminates against contemporary forms of art.

With the recording of dreams, copyright requirements align more closely with UK law's criteria for speech and conversation. The Beijing Treaty offers a potential workaround for streaming dreams by allowing protection for unfixed works. The critical question about copyrighting dreams hinges on whether dreams can be intentionally created or if they are mere byproducts of biological functions. The complexity deepens when considering the creation process of artificial intelligence, which operates not on a "force of nature" basis but as a set of neural networks with arousal indicators.

A question of intent and originality

"The dream is of a first-person experience of being trapped on a radioactive planet. The only means off the planet to spaceship waiting in orbit is to catch an elusive red bird with teleportation powers. The twist that is revealed later is that the planet was the Earth all along. Before broadcasting the dream, it is reviewed and then published without editing."

American lawyer David Nimmer has examined intent as a crucial element in copyright law for determining the originality of a work. His central argument emphasizes that an author must intend to produce an original work. The question of intent in determining the originality of dreams arises from the nature of dreams. As argued in the case of Kelley, natural forces cannot be authors of original works. Dreams, however, originate from a human being, unlike wildflowers, and cannot be considered a similar force of nature.

The critical question is whether the creation, to be considered original, must be intentionally and consciously produced, or if a mere accidental byproduct of a chemical process in the brain can be granted protection. A similar case that sparked the discussion of intent involved the reconstruction of the Dead Sea Scrolls by archaeologist Elisha Qimron and the unauthorized publication of these reconstructions by an American publisher.

In the reconstruction of the scrolls, understanding linguistics of the ancient language and ancient Jewish laws was crucial to supplement missing or damaged fragments. Professor Elisha Qimron joined the reconstruction team as a specialist in these fields. The court concluded that by publishing the scrolls, the intellectual property rights of Elisha Qimron were infringed. The Israeli Supreme Court ruled in favor of Elisha Qimron, emphasizing the intent in creating an original work.

Traditionally, there has not been a need to distinguish intent from the act itself. A canvas needs painting, a book needs writing, a camera needs to be operated—all deliberate acts. The question arises: do we intend to dream, or is deciding to publish a dream enough to prove intent in creating an original work?

The question of the Dead Sea Scrolls dealt with a similar matter. The Israeli Supreme Court ruled that the lack of intent is not sufficient proof to discard the originality of the work. The same question arises when determining the intent in using artificial intelligence to create a work. Is the intent exemplified by the directions of the prompt, or is it exemplified by filling in the gaps with digital knowledge?

There is strong evidence that dreams are synthesized by the forebrain, 'comparing information generated in specific brain stem circuits with information stored in memory.' While dreaming, an individual is not in control of the content of their dreams, unlike the case of the Dead Sea Scrolls. An individual would not be able to actively participate in the creation of dreams. Unlike Prof. Qimron, an individual would not be able to choose not to dream about the twist of the story.

The originality requirement is a defining characteristic of copyright law, consisting of two elements: independent work and creativity. Some argue that creativity is an obsolete requirement, especially as there is no legal definition for creativity. The Israeli Supreme Court ruled in favor of Prof. Qimron, indirectly recognizing the independent work aspect as a sufficient characteristic of originality.

It is not yet possible to technologically or scientifically measure, quantify, and allocate with sufficient precision the working processes of the human brain. While ample research suggests that what we consider conscious choice is a result of uncontrolled chemical reactions within our brain, no different from chemical reactions occurring in a dream state, there is currently no agreed-upon definition of what constitutes creativity or how to measure it with sufficient accuracy. The question of whether dreams could be original remains open. Neural networks, however, could provide us with quantifiable data to understand the process of creativity of an artificial intelligence. These understandings might have spill-over effect in measuring the processes within human brain.

As seen in the case of the Dead Sea Scrolls, circumstances of individual cases can provide sufficient arguments in favor of recognizing originality based on the work itself. In a dream scenario like B, where the individual decides to broadcast the dream after reviewing it, the presence of the review process itself could be sufficient to merit that the work is original.

A question of expression

"First person experience of being trapped on a radioactive planet. The only means off the planet to spaceship waiting in orbit is to catch an elusive red bird with teleportation powers. The twist that is revealed later is that the planet was the Earth all along. Before publishing the dream, the twist part is removed."

In the judgement C-310/17 the European Court of Justice stipulated two critical requirements to be filled for a “work” to be protected as an intellectual property – a) “it needs to be original in the sense that is the authors own intellectual work “and b) only an expression of this intellectual creation can be considered “work”. The court also, however, mentions that the work “must be expressed in a manner which makes it identifiable with sufficient precision and objectivity, even though that expression is not necessarily in permanent form.“

The question of expression is closely tied to the question of fixation as original works are defined by their fixation in medium of expression. Similarly, the ECJ stated that only an expression of intellectual creation can be considered work as mentioned above. The expression should be the differentiation from an idea. Idea is manifested through its expression. In a case of dreams, idea and its expression might not be as easily differentiated. It also calls to question the notion of the ECJ that the expression needs to be of intellectual creation. As mentioned in the previous chapter, there are no agreed upon definitions of creativity. It becomes more complicated to try and define what is “intellectual”. In the case of the Dead Sea Scrolls prof. Qimrons reliance on his expertise in the matters of ancient languages and Jewish laws in recreating the ancient texts can be considered a result of “intelligence” as we culturally understand it. It is difficult to argue against a proactive decision-making process that is based on previous knowledge in a specific discipline as not to be an intellectual work. Yet, an artist creation of spontaneous improvisation performance of modern dance or painting that rely on artists’ intuition of the base-line knowledge creates more challenges to be called “intelligent”. It would more likely be based on subjective understandings of what is an intellectual work and therefore the second requirement for dreams in order to be expressed might not be filled

The Beijing Treaty aims to fill exactly these gaps as it recognizes the artistic and creative value of unfixed and spontaneous performances. This creates a separation between fixation and an expression. The Article 6 (i) of the Beijing Treaty states that performers shall enjoy the exclusive right of authorship of the broadcasting and communicating their unfixed performances. Interpretation of the Article 6 (i) could showcase that the fixation is not assumed as part of the expression. Expression can be separate from the fixation form, especially in the case of dance in where the artist uses his/her body as a tool of the expression of the idea.

Similar analogues can be drawn in the expression requirement for the dream. The Beijing Treaty gives protection for unfixed improvised works as well as recognizes singers, musicians, dancers, and actors as subjects of audiovisual performers. Indirectly the Beijing Treaty accepts human body as a tool of expression for intellectual work. Therefore, one could interpret this acceptance in an expansive way and conclude that brain’s ability to create dreams is similarly using the tool of a human body in expressing intellectual work.

The question of intelligent work in the case of artificial intelligence hinges on our acknowledgment of its technical capabilities as sufficiently cognitive and aware to be considered intelligent in terms of independent decision-making. In its current state, artificial intelligence acts more as a helping tool, relying on its database of digital knowledge and following human directions. It lacks the capacity for independent initiation of tasks.


In his article about fixation Evan Brown writes “For most of the history, fixation has not been an issue. It was simply an undifferentiated part of the authorship process.” This idea of obsoleteness can be expanded to many aspects of copyright law when considering the copyrightability of dreams.  

The evolution of technology and our growing knowledge of human brain processes continuously challenge the principles of the copyright law. From a purely cultural perspective, the basis of our copyright law is rooted in Christianity and carries in itself ideas of higher creation, purpose and human soul. The 19th century core principles of copyright law mirror the Christian God creation story channeled through a human agent. It makes any other agent incapable of creation as access to creation can only be achieved through immortal soul. This view is more in line with the naturalist origin of the copyright law in where copyright law stems from authors natural right to the fruits of their labor. From this view, artificial intelligence, as an entity not capable of having soul, can never truly create.

Copyrighting dreams creates unique challenges in the basic principles of intellectual property. The requirements of fixation, expression and originality need to be reconsidered. As was seen in the Beijing Treaty, the evolvement of the digital age has already helped expanse the requirements for an intellectual work in the case of fixation by recognizing and giving protection for unfixed works. The question in need to be solved for dreams to be fully protected would be the question of creativity and originality.