Vai al contenuto

Part two. Who is the author today?

In the first part of this book, we explored the concept of hybrid author, analyzing how contemporary creativity arises from the synergistic interaction between human and artificial intelligence. We have seen how the role of the author is not diminished by the use of AI, but evolves into a new form of creative expression, where the artist's choices—from the initial conception of the work to the final selection—remain central to the creative process.

This evolution of the authorial role inevitably raises questions on the legal level. How does this new creative paradigm translate into the world of law? How are legal systems reacting to works born from the collaboration between human mind and artificial intelligence?

Courts and authorities in different countries have already had to face these questions, producing a series of decisions that are outlining the contours of this new reality. Even if with a certain delay, but as is in the nature of things, legislators are beginning to propose specific regulations for this emerging phenomenon.

In the following chapters, we will analyze some particularly significant rulings, regulations, and regulatory proposals, selected to offer both a geographical and temporal perspective. The cases span from the United States to China, passing through Europe, and extend temporally from the historic Sarony ruling of 1884 to the very recent Italian bill in the approval phase. The choice of these cases is not random: individually each of them tells a specific aspect of the question, but, seen together, they reveal a surprisingly coherent common thread in the approach to authorship. From the first debates on photography in the 19th century to the current controversies on AI-generated images, it will emerge how different jurisdictions, in different eras and cultures, have developed similar approaches in recognizing and protecting new forms of creativity, always keeping the role of the human author at the center.

The decisions of the courts and the new regulations are already concretely shaping the future of digital creativity, influencing the possibilities available to artists, creatives, and professionals who use AI in their work.

Chapter 4. What rulings and regulations say

As we have already extensively explained, the spread of generative artificial intelligences raises a crucial question: who should we consider the author of a work created with the assistance of AI? This question has significant economic and creative implications in the field of creativity and intellectual property.

Generative AI, capable of producing high-quality content, makes the distinction between human creator and technological tool less clear-cut. The question is whether a work generated by an AI can be protected by copyright and, if so, who holds the rights. It is also a matter of understanding whether this tool, conceived when creativity was exclusively human, is still valid or is superseded by what technology is proposing to us.

To analyze the question, after a first look at the guidelines of the U.S. Copyright Office, the first to have to confront the arrival of AI, we will examine both historical and current cases. The Napoléon Sarony vs Burrow-Giles case of 1884, which established the existence of copyright on photographs, offers a precedent on how the law has adapted to new forms of creative expression.

In the current context, various judicial cases are contributing to defining the debate on authorship in the AI era. These are cases that have led to identifying the emerging criteria for determining authorship in this new technological context, such as "substantial human creative contribution" and "creative control over the final result."

The first to move to try to put rules on these aspects was the U.S. Copyright Office; after all, it is in the USA that the first requests for copyright for works generated with AI arrived immediately, even reaching cases that have become famous in the meantime. The guidelines that came into force on March 16, 2023, represent an attempt to guide the practice of authors, adapting the traditional concept of copyright to the new challenges posed by artificial intelligence. [Copyright Office, Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence, March 16, 2023]

According to this important document, the criteria for evaluating the possibility of copyright on works created with the assistance of AI are:

  • Substantial human creative contribution beyond "de minimis": the work must show a significant creative input from the human user; the human contribution must exceed a minimum threshold of creative relevance.
  • Creative control over the final result: the user must have exercised a determining control over the AI's output.
  • Selection and arrangement of AI-generated material: the choice and organization of content produced by the AI must reflect human creativity.
  • Substantial modification of AI-generated material: the modifications made by the user to the AI's result must be significant.
  • Use of AI as an assistive tool: the AI must be used as a tool serving human creativity, not as an autonomous creator.
  • Formation of "traditional elements of authorship": the work must present characteristics typical of human authorship.

These guidelines reflect a balanced approach that recognizes the creative potential of AI while maintaining the emphasis on human contribution. The Copyright Office evaluates each request individually, considering to what extent the work satisfies these criteria. The most relevant aspect of these guidelines is the focus on the active role of the human user in the creative process. It is not sufficient for the work to simply be generated by a human user with an AI; there must be a substantial and demonstrable human contribution.

This approach raises some questions. For example, how is "substantial contribution" quantified and how is the existence of "creative control" verified? The definition of these terms could vary significantly depending on the context and the AI technology used.

Furthermore, the emphasis on modification and selection of AI-generated material suggests that the creative process is not limited to initial generation but also includes the post-production and selection phase.

4.2 Significant case studies

The evolution of the concept of author in the digital age, and in particular in the era of generative AI, reveals a surprising consistency across different jurisdictions and technologies. Analyzing some key rulings, a common thread emerges that, while adapting to different cultural, legal, and technological contexts, addresses fundamental questions about authorship in a uniform way. These case law cases offer valuable insights on the current challenges related to the authorship of works generated with the assistance of advanced technologies, including AI.

In this section, we will examine five representative cases.

Napoléon Sarony vs Burrow-Giles (1884): this historic ruling, although not directly concerning AI, established fundamental principles on the recognition of authorship in works created with the assistance of technologies (in this case, photography). Obviously it is here for the parallels it offers with current discussions on generative AI.

Thaler vs Copyright Office (2023): a recent case, the request, perhaps it would be better to call it the provocation, by researcher and inventor Stephen Thaler who asked that copyright be recognized for a work generated autonomously by an AI specifically built by him.

Li vs Liu (2023): a recent case from China that directly addresses the question of authorship of images generated with AI. This ruling provides valuable insights on how courts are beginning to interpret and apply copyright laws in the context of generative AI.

Zarya of the Dawn (2022): this case, involving the U.S. Copyright Office, explores the limits of authorship in a work composed of text and AI-generated images. Despite the refusal to grant copyright on the images, it is fully in line with other rulings.

Biancheri vs RAI (2023): an Italian ruling that addresses the question of authorship in works created with algorithms. This case shows how the principles established for advanced technologies can be applied consistently to generative AI.

Through the analysis of these cases, we will try to trace a complete picture of the current legal challenges and interpretations regarding authorship in the era of generative AI, highlighting how, despite geographical and temporal differences, a surprisingly coherent approach emerges in evaluating human creative contribution.

In the most recent cases, I have tried to show also the technical and temporal steps followed by the judges, which has led to having more tables, even of considerable lengths, for each case. To facilitate reading, I have preferred to separate these tables from the narrative of the events. The tables have been moved to the Appendices section at the end of the volume.

4.2.1. Napoléon Sarony vs Burrow-Giles (1884)

The Napoléon Sarony vs Burrow-Giles case, although dating back to 1884, offers interesting parallels with current discussions on the authorship of works generated with the assistance of AI. [Burrow-Giles Lithographic Co. v. Sarony, Wikipedia] This historic ruling laid the foundations for the recognition of photography as an art form worthy of copyright protection, addressing questions that resonate today in the debate on generative AI.

Napoléon Sarony, a famous portrait photographer of the time, sued the Burrow-Giles Lithographic Company for reproducing without authorization his portrait of Oscar Wilde from 1882. At the time, photographs were not protected by copyright, as it was believed that this technique allowed only a mere mechanical reproduction of reality, devoid of human creativity, leaving the photographer the trivial task of taking the shot.

The Supreme Court of the United States, however, after careful examination ruled in favor of Sarony, establishing that photographs could be considered works of art protected by copyright.

Napoleon Sarony – Portrait No. 18 of Oscar Wilde

The key motivation of the ruling reads:

posing the said Oscar Wilde in front of the camera, selecting and arranging the costume, draperies, and other various accessories in said photograph, arranging the subject so as to present graceful outlines, arranging and disposing the light and shade, suggesting and evoking the desired expression, and from such disposition, arrangement, or representation, made entirely by the plaintiff [that is, Sarony Ed.], produced the photograph in suit.

This detailed description of Sarony's creative actions highlights how the photographic process was not to be considered even then a mere mechanical reproduction, but an activity that required artistic and creative choices.

The parallelism with generative AI is, I believe, evident. Today, as then, we are faced with a technology that seems able to act automatically. And, as then, the central question concerns the creative role of the human being in the process. The choices of a generative AI user can be seen as analogous to Sarony's creative choices in composing his photograph.

This historic ruling reminds us that the introduction of new technologies into the creative process does not necessarily diminish the role of the human author. Rather, it shifts the focus of creativity to new aspects of the process. In the case of generative AI, as for photography in the 19th century, the challenge (both on the legal and conceptual level) lies in recognizing and valuing these new forms of creative input.

The Sarony case ruling offers several surprisingly current insights for understanding the relationship between creativity and artificial intelligence. First, it helps us look beyond technology in the creative process. Just as Sarony demonstrated that photography was not reduced to the simple click of a button, today artists working with AI emphasize how their creative process is much more complex than typing a prompt and pressing enter.

A second crucial aspect concerns the value of creative choices. The Court, in the Sarony case, recognized the fundamental importance of artistic decisions in composition, lighting, and posing of the photographic subject. Similarly, today we can consider how an author's choices in using AI—from selecting the most appropriate models to the complex refinement of prompts—can be determining in defining the authorship of a work.

The Sarony affair also teaches us something about the law's ability to adapt to innovation. The legal system then demonstrated that it could evolve to understand and regulate new technologies. Today we face a similar challenge with artificial intelligence, and this historical lesson can guide us in addressing the new legal questions that emerge.

Finally, that historic case contributed significantly to redefining the very concept of originality in works of art. It is interesting to note how, in the AI era, we find ourselves once again having to rethink what it means to create something original, in a completely new technological context but with surprisingly similar questions.

The practical result of this lawsuit was that for the first time a photographer was able to add this inscription under his photo. [Simone Aliprandi also talks about it on his YouTube channel: https://youtu.be/fbxNIB6Y5PQ]

Caption appended to the photo of Oscar Wilde after the ruling

The Thaler vs Copyright Office case, decided by the U.S. District Court for the District of Columbia in August 2023, represents a milestone in jurisprudence regarding works generated with the assistance of AI. [United States District Court, District of Columbia, Thaler v. Perlmutter, 2023] This ruling directly addresses the question of authorship and copyright protection for images created using generative AI systems without human intervention. The case concerns inventor Stephen Thaler, who attempted to register a visual artwork entitled A Recent Entrance to Paradise with the U.S. Copyright Office. Thaler claimed that the work had been created autonomously by an artificial intelligence system called Creativity Machine (of which he is the owner and which he himself designed), without any human intervention in the creative process.

The recent court ruling established a fundamental precedent in the debate on artificial intelligence and copyright. At the center of the controversy, we find the Copyright Office's decision to deny the registration of a work, based on a fundamental principle: the absence of direct human creative contribution. This decision was not taken lightly but represents the result of careful legal analysis.

The court, confirming the Copyright Office's position, reaffirmed a key concept in U.S. law: copyright protection is reserved exclusively for works of human creation. To reach this conclusion, the judges conducted an in-depth analysis, examining not only the text of the copyright law but also its legislative history and case law precedents. This methodical approach allowed highlighting how human authorship is not a simple technical detail but a fundamental requirement of the copyright protection system.

A particularly interesting aspect of the ruling concerns the distinction between creative tools and creators. The court recognized that artificial intelligences can certainly be used as tools in the creative process—just like brushes, cameras, or editing software—but clarified that they cannot be considered "authors" in the legal sense of the term. This distinction is crucial for understanding the role of AI in the contemporary creative process.

Looking to the future, the ruling demonstrates a clear awareness of the challenges that the increasing use of artificial intelligence will bring to the artistic and creative field. However, while recognizing this technological evolution, the court maintains a firm position: human authorship remains, at least for now, an essential element for copyright protection. This decision does not close the doors to innovation but establishes clear boundaries for the future development of the relationship between human creativity and artificial intelligence.

This ruling offers important insights for creators who use generative AI. First, it underlines the importance of human intervention in the creative process to obtain copyright protection. Furthermore, it clarifies that, currently, works generated entirely by AI without significant human input are not "copyrightable" in the United States. Finally, it suggests that human involvement in the creative process will need to be documented and substantial to obtain copyright protection for works involving the use of AI.

The Thaler vs Copyright Office case represents an important precedent in the recognition of AI-generated works and offers valuable guidance for courts around the world that will find themselves facing similar cases and should be considered consistent with the Copyright Office guidelines, since the rejection of Thaler's request is based on the lack of insufficient creative input from a human user of the machine.

It also demonstrates how the legal system is trying to balance technological innovation with traditional copyright principles.

For a detailed technical analysis of the points examined by the court and a complete chronology of the actions taken by Thaler in the creative and legal process, see the appendix in Appendix A.

4.2.3 Li vs Liu (2023)

The Li vs Liu case, decided by the Beijing Internet Court in November 2023, directly addresses the question of authorship and copyright protection for images created using generative AI tools. [Beijing Internet Court, The Li v. Liu Case, Italian translation available on dirittocinese.com]

Briefly, Li generated an image using Stable Diffusion (specifically, the Automatic1111 interface) and shared it on the Chinese social network Little Red Book. Li subsequently discovered that Liu had used the image, removing the watermark, to accompany a text published on another social network. Li then sued Liu for copyright infringement.

The court's decision was clear: the use of artificial intelligence tools does not prevent the recognition of copyright, provided there is significant human creative contribution. To reach this conclusion, the court conducted an in-depth analysis of Li's creative process, examining in detail their artistic choices: from the selection of the most appropriate AI software, to the choice of models, to the formulation of textual prompts and the setting of technical parameters.

Particularly interesting is the way in which the court framed artificial intelligence in the creative context: it equated it to other artistic tools, such as cameras or editing software. This perspective underlines how it is the artist's creative control, not the tool itself, that determines the authorship of the work. It is the artist who, through their choices and their vision, transforms the tool into a means of artistic expression.

In the specific case, the court also addressed a practical question of copyright infringement, establishing that Liu, by using Li's image without authorization and removing the watermark, effectively violated the artist's copyright. This decision strengthens the legal protection of works created with the assistance of AI, equating them to those created with more traditional means.

A final significant aspect of the ruling concerns transparency in the use of AI. While encouraging artists to be open about the use of artificial intelligence tools in their creative process, the court did not consider it necessary to impose a legal obligation of disclosure. This decision leaves artists free to decide how to communicate their creative methods, while emphasizing the importance of transparency in dialogue with the public.

This ruling offers important insights for creators who use generative AI, as it highlights the importance of documenting the creative process, including the choices made during the use of AI tools. It also underlines that authorship can be recognized even when AI tools are used, provided there is significant creative contribution from the user. And furthermore, it suggests that copyright protection extends to AI-generated works, opening new possibilities for creators in this field.

For a detailed analysis of the points examined by the court and a complete chronology of the actions taken by Li in the creative process, see the appendix in Appendix B.

4.2.4 Zarya of the Dawn (2022)

The Zarya of the Dawn case represents a turning point in the discussion on the authorship of works generated with the assistance of AI in the United States. [Richard Lawler, The US Copyright Office says you can't copyright Midjourney AI-generated images, The Verge, February 23, 2023] This case, which involved the U.S. Copyright Office in September 2022, raised crucial questions about the recognition of copyright for works created using generative AI tools.

Kris Kashtanova, a prominent figure in the world of creative AI, submitted a copyright application to the Copyright Office for Zarya of the Dawn, a comic story created using Midjourney. [The work is available on the author's website: https://www.kris.art/portfolio-2/project-one-ephncjamy8]

The Kashtanova case represents an interesting precedent in the complex relationship between artificial intelligence and copyright. The affair took a significant turn when the Copyright Office, after having initially granted registration of the work, discovered that AI tools had been used—a detail not declared in the original application. This discovery led to the withdrawal of the registration and the initiation of a new, in-depth analysis process.

This affair highlighted how complex it is to evaluate works created with the assistance of artificial intelligence and how fundamental transparency is in the copyright registration process. The Copyright Office conducted a meticulous analysis of Kashtanova's creative process, examining in detail their role in the various phases of creation of the work. This analytical approach allowed establishing a fundamental distinction between the different elements of the work.

Particularly interesting is how the Copyright Office made this distinction. The text was recognized as entirely Kashtanova's work, therefore worthy of full protection. The compilation and arrangement work, which included both the text and the layout of the images, was also judged worthy of copyright protection. However, the individual images generated through Midjourney received different treatment: they were not deemed eligible for copyright protection because, according to the Copyright Office, Kashtanova did not exercise sufficient control of the creative process.

This decision is based on a careful evaluation of the role of AI in artistic creation. According to the Copyright Office, it was Midjourney, and not Kashtanova, that determined the fundamental expressive elements of the images. This placed a significant limit on the recognition of human authorship in the work. Even the modifications made by Kashtanova to the images after their generation were considered too marginal to reach the threshold of creativity necessary for copyright protection.

The Zarya of the Dawn case marked an important passage in the definition of the relationships between artificial intelligence and copyright, also because it was the first case of this kind with considerable media resonance. The final decision created a significant precedent that deserves careful reflection.

This decision highlighted several fundamental aspects for those who work in the creative field using artificial intelligence. First, the importance of transparency: declaring the use of AI tools in one's creative process is not only an ethical question but becomes a crucial element in the copyright registration process. The Zarya of the Dawn affair demonstrates how failure to declare can lead to significant complications and potential invalidation of the protection obtained.

Another central aspect that emerged from this case concerns the way in which the Copyright Office evaluates human contribution in works created with the assistance of AI. It is not enough to simply use an artificial intelligence tool: it is necessary to demonstrate significant creative control and substantial artistic input in the creation process. This establishes an important precedent on the limits of copyright for elements generated primarily by AI, emphasizing how legal protection is closely linked to the presence of a documentable and significant human creative contribution.

The case also highlighted the complexity in evaluating hybrid works, those that combine human-created and artificial intelligence-created elements. Zarya of the Dawn represents a perfect example of this new category of works, where the line between human and artificial contribution is not always easily distinguishable. This complexity pushes towards a broader reflection on the very evolution of the concept of authorship in the era of generative AI, raising fundamental questions about what it means to be an "author" in a context where technology plays an increasingly significant role in the creative process.

For artists working with AI, this case offers valuable practical guidance. It becomes fundamental not only to accurately document one's creative process but also to understand the current limits of copyright in relation to works generated with the assistance of artificial intelligence. This awareness allows structuring one's work in order to maximize legal protection of one's creations, while maintaining an innovative approach to the use of new technologies.

Having been perhaps the first case of a copyright request for works generated also with the assistance of one or more generative AI, knowledge on how to handle the matter was probably lacking on both sides, and perhaps this led to non-optimal decisions.

Subsequently, Kris Kashtanova submitted a new application to the Copyright Office [Morrison Foerster, Registration Cover Letter] for a further work, entitled Rose Enigma. [https://www.kris.art/portfolio-2/rose-enigma] This time developed with Stable Diffusion and a set of other software, in the letter accompanying the request all the choices made in function of the control of the final image are listed, in particular the use of ControlNet Depth, a Stable Diffusion extension, with a reference sketch. The choice of Stable Diffusion because among the various generative AI for images it has always been the one that allows the best control of the process.

At the editorial closing date of this book, there is no indication that the Copyright Office has yet responded.

For a detailed analysis of the points examined by the Copyright Office and a complete chronology of the actions taken by Kashtanova in the creative process, see the appendix in Appendix C.

4.2.5 Biancheri vs RAI (2023)

The Biancheri vs RAI case, concluded with order 1107 of the Court of Cassation on January 16, 2023, represents an important precedent in the Italian legal landscape regarding the authorship of works created with advanced software. [Court of Cassation, Order no. 1107/2023]

Architect Chiara Biancheri sued RAI for having used, without authorization, her graphic work The scent of the night as scenography for the Sanremo Festival 2016. The work had been created using Apophysis, software for generating fractals. Although the ruling does not directly deal with artificial intelligence, the case presents analogies with the questions raised by the use of AI tools for the creation of artistic works due to the use of creative tools based on algorithms inserted as part of a complex creative workflow.

The analysis of the case reveals several important aspects in the legal evaluation. Through all three levels of judgment—from the Genoa Court to the Court of Cassation—Biancheri's work was constantly recognized as worthy of copyright protection. The value given by RAI in the presentation of the event also contributed to this.

The courts confirmed the creative character of the work, demonstrating that the use of advanced software does not diminish the artistic value of the creation; the fact that the software generates images through algorithms was not considered an obstacle to the recognition of Biancheri's authorship. This latter aspect can also be seen as an implicit parallel with the historic Sarony vs Burrow-Giles case. The courts placed particular emphasis on the need to evaluate the degree of creative elaboration of the artist, even when advanced technological tools are used.

This ruling has important implications for the future of digital creativity and generative AI. First, it establishes a fundamental principle: the use of advanced software does not automatically exclude the recognition of creativity and human authorship. The courts gave particular weight to the process of artistic elaboration, a principle that could also be applied to works created with generative AI. The importance of a case-by-case analysis of the artist's creative contribution also emerges, avoiding superficial generalizations.

The case also opens new perspectives for the future: the recognition of authorship in these circumstances could favor more inclusive interpretations for works created with emerging technologies, including AI. Although Apophysis is not a generative artificial intelligence, this case provides a valuable precedent on how the legal system can approach works created with software that incorporates elements of automation or learning.

The Court of Cassation rejected RAI's appeal, confirming the decisions of the previous levels, ordering it to pay legal costs and confirmed the compensation of forty thousand euros in favor of Biancheri.

Although the Court declared RAI's appeal inadmissible on procedural grounds regarding the question of the image generated with an algorithm, it nevertheless provided important considerations:

The counterpart's admission of having used software to generate the image is certainly not sufficient for this purpose, a circumstance which, as the appellant herself admits, is in any case compatible with the elaboration of a work of genius with a rate of creativity that should only be scrutinized with greater rigor, if, as happened in the specific case, RAI did not ask the merit judges to reject the claim for that reason.

The Court further emphasized:

And in fact a finding of fact would have been necessary to verify whether and to what extent the use of the tool had absorbed the creative elaboration of the artist who had used it.

These observations highlight the need for an in-depth analysis of the creative process, even when advanced technological tools are involved.

The ruling also includes a statement from the Court of Appeal, whose decision was confirmed by the Cassation:

In this case, the Court of Appeal observed that the work is creative when it expresses an original idea, coming only from the inspiration of its author and confirmed the evaluation expressed by the first instance judge, arguing that the image was not a simple reproduction of a flower, but involved a true re-elaboration, therefore worthy of copyright protection for its creative character.

This ruling aligns with the previously analyzed cases, suggesting that an image generated with the assistance of advanced technologies can be considered creative and enjoy copyright protection, provided there is significant creative effort from the user. In a sense, the ruling suggests that the use of advanced technological tools is not in itself sufficient to deny the creativity of a work. Rather, it is necessary to carefully evaluate the user's creative contribution in the process of creating the work. For a detailed analysis of the points examined by the various courts and a chronology of the legal proceeding, see the appendix in Appendix D.

4.3 The Italian proposal: the AI bill

In May 2024, the Italian government presented a bill on artificial intelligence that, among the various aspects addressed, directly addresses the question of authorship of works created with the assistance of AI. [DDL 1146, Provisions and delegation to the Government on artificial intelligence, available on senato.it] Article 24 of the bill proposes a significant modification to the copyright law (L. 633/1941), explicitly inserting works generated with the assistance of AI among those protected by copyright, provided they constitute "the result of the intellectual work of the author."

It is interesting to note how this proposal is grafted onto a regulatory basis that already places creativity at the center of copyright protection. In fact, article 1 of law 633/41 protects "works of creative character," placing creativity as a fundamental requirement for protection. The modification proposed by the bill aligns perfectly with this principle, extending its application to works created with the assistance of AI without distorting its essence.

This proposal represents a further step in the evolution of copyright with respect to new technologies. As we have seen in the Sarony vs Burrow-Giles case of 1884, the legal system has already had to face in the past the challenge of adapting to new creative tools. Then it was photography, today it is artificial intelligence. In both cases, the focus is not on the tool used but on the creative contribution of the human author.

The parallel with the Sarony case comes spontaneously. Just as the American Supreme Court recognized that the use of the camera did not preclude the photographer's creativity, so the bill recognizes that the use of AI does not preclude the author's creativity. In both cases, it is the person's intellectual work—not the tool used—that determines the protectability of the work. This proposal aligns perfectly with the orientation that emerged in the most recent cases discussed in the previous paragraphs. The guidelines of the American Copyright Office, the Chinese Li vs Liu ruling, and the Italian Biancheri vs RAI case all highlighted the importance of human creative contribution as a determining element for the recognition of authorship.

The proposed modification is significant because it does not limit itself to interpreting existing law but explicitly updates it to include the new creative realities. The inclusion of a direct reference to artificial intelligence in the copyright law would represent a strong legitimization for artists and creatives who use these tools, offering them a clear legal basis for the protection of their works.

The wording chosen by the legislator is particularly interesting. The proposal maintains the emphasis on the "intellectual work of the author" as a fundamental criterion, consistent with the principles that emerged in international jurisprudence. This balanced approach recognizes the new possibilities offered by AI without compromising the fundamental principle that links copyright protection to human creativity.

Article 24 of the bill also aligns with the vision of the hybrid author discussed in the first part of this book. It implicitly recognizes that the use of AI does not diminish the creative role of the human author but can be an integral part of the contemporary creative process. This vision reflects the reality of many creatives who today use AI as a tool to expand their expressive possibilities.

If approved, this legislative modification could have a significant impact on the Italian creative landscape. Not only would it offer greater legal certainty to authors who use AI, but it could also encourage innovation and experimentation with these new technologies, knowing that their creative work will be adequately protected by law.

However, the bill presents some critical issues that deserve attention. In particular, article 12 limits the use of AI in intellectual professions "exclusively to exercise instrumental and support activities," requiring the "prevalence of intellectual work." This restriction seems to arise from a limited understanding of how AI can integrate into the creative and intellectual process, not only as an accessory tool but as an integral part of the process itself. Such limitation appears in contradiction with the spirit of article 24, which instead recognizes the creative value of the use of AI.

Article 23, which imposes the obligation to identify AI-generated content through markings or other means, although aimed at transparency, risks creating unjustified discrimination towards works created with the assistance of AI. Furthermore, it seems in contradiction with article 24: while the latter recognizes the parity of works created with AI (provided they are the fruit of intellectual work), article 23 "marks" them as different, potentially influencing their perception and acceptance.

In conclusion, article 24 of the bill represents an important step forward in the legal recognition of new forms of creativity, even if other parts of the text would need revision to fully align with this innovative vision. While maintaining the fundamental principles of copyright, the proposal adapts to new technological realities, offering a clear regulatory framework for the protection of works created with the assistance of AI. This approach, if perfected by eliminating the inconsistencies highlighted, could serve as a model for other jurisdictions facing similar challenges in the era of artificial intelligence.