A global group of creativity researchers has issued a strong call to action: artificial intelligence is transforming the domain of creativity at an unprecedented pace. The challenge today? Deciding which future we want to pursue.

Multidisciplinary experts from the AI Task Force of the International Society for the Study of Creativity and Innovation (ISSCI) have presented, in a scientific paper – “Cyber-Creativity: A Decalogue of Research Challenges” -, a decalogue of dimensions related to cyber-creativity, associating each of them with a set of research questions and related challenges. The aim is not, of course, to provide answers, but to orient the conversation around the forms of collaboration between humans and AI in creative work. In the background lies a key question: how can we ensure that AI becomes our most powerful creative partner without eroding human agency, culture, and originality?

Is creativity no longer exclusively human?

Before exploring the ten points, it is important to clarify the definition of the key concept introduced in the article. For centuries, creativity has been considered an exclusively human domain, yet generative AI (including language models as well as image and music generators) now produces outputs that can compete with those of professionals.

We therefore need a new theoretical framework to guide our understanding and use of a field that, according to the experts, should become a structured, interdisciplinary, and ongoing area of research.

“Cyber-creativity” is defined as any creative process involving collaboration between humans and AI, across a spectrum ranging from human-driven to machine-driven processes: not in a relationship of opposition (humans vs. machines), but from the perspective of understanding the relationship between the two.

Exploring possible futures as a research tool

With an unconventional approach inspired by futures and foresight practices, the article opens by presenting two contrasting scenarios for 2035.

The first is a utopian future, where AI enhances human creativity rather than replacing it, where ethical frameworks ensure fairness, transparency, and bias control, and where artists are protected by equitable compensation systems. In this scenario, education is personalized and oriented toward creativity, and soft skills are supported by a technological ecosystem that amplifies human capabilities. Sustainability improves through AI-assisted innovation, and human imagination remains central. It is a system of amplification, not suppression, of the creativity inherent in human nature.

This is contrasted with a more dystopian future, where technological monopolies control creative production, stripping artists of ownership over their work. Culture becomes flat and homogenized, driven by opaque and pervasive algorithms. Human creators are marginalized by low-cost AI alternatives, and creative diversity is significantly reduced. Education, too, becomes automated and superficial: in this scenario, AI acts as an “extractor” rather than a partner.

The “Decalogue”: 10 urgent challenges for the future of cyber-creativity

According to the authors of the paper, it is not too late to steer humanity toward the positive scenario, provided that we focus on ten priorities.

1) Building a new theory of creativity

We need to develop a theoretical model that frames the characteristics of human creativity, artificial creativity, and their interaction. It is not necessary to argue that AI creativity is identical to (or even comparable with) human creativity, just as it would not be meaningful to equate creative behavior in animals with the unpredictable evolution of cosmic matter. All forms of creativity are distinct, yet coexist within a unified framework. The challenge, therefore, is to study this complex phenomenon through an interdisciplinary approach, recognizing its dynamic nature and rapid evolution, without focusing solely on the final output.

2) Understanding cultural impact

Creativity has a social dimension: from this perspective, it is essential to recognize that human creativity is simultaneously psychological, social, and material. This implies exploring how interactions between humans and AI may enhance certain cognitive capacities while attenuating others, and how variations in human expertise influence the roles assigned to AI.

Over time, system dynamics may evolve, potentially reaching tipping points that signal radical change. It becomes important to explore and anticipate how human–AI creative interaction may reshape cultural norms and power structures, considering AI’s dual role both as a “cultural artifact,” shaped by societal values, and as an agent that reconfigures the landscape of creative production itself.

3) Rethinking the creative process

How does human–AI co-creation actually unfold? While current AI models tend to reduce variability in creative performance, variability is a defining feature of human creativity and must remain central in the process.

Moreover, an optimal co-creative process is likely one in which humans retain control over AI-generated outputs, in order to mitigate non-human biases.

AI can be particularly useful in the problem-definition phase, enabling deeper exploration at the outset. Beyond serving as a tool for gathering relevant information, it can also function as a source of inspiration. However, the authors emphasize the importance of an incubation period for the human counterpart, since AI systems tend to move quickly toward conclusions.

This suggests the need to introduce forms of “friction” that counteract this linearity, within a process that is, by definition, initially divergent before becoming convergent. A productive tension must therefore be maintained between exploration, intuition, and inspiration—areas where systems should be designed to stimulate questioning and openness rather than inhibit the “fermentation of ideas.” AI, when aligned with creativity, supports experimentation (and possibly failure), while remaining an ally in achieving efficiency during the implementation phase.

4) Redefining the creative “agent”

How should authorship and originality be defined in work produced with AI support? Further research is needed to delineate the boundaries of this collaborative role, preventing AI from becoming a controlling force or subtly undermining human agency.

AI should act as a facilitator, not a substitute for human ingenuity, relying on systems that ensure transparency, user control, and reliable performance. Key factors include data quality, algorithmic bias, and the balance between automation and human intervention, all of which are crucial for designing interfaces that are both effective and ethically responsible.

There is also a significant dimension related to well-being: human happiness and creativity are intrinsically linked, and preserving psychological well-being is essential in a context where digital technologies do not always produce positive effects. Concerns related to ethics and AI reliability are part of this landscape: regular audits of digital well-being for individuals and organizations should become standard practice.

5) Studying co-creative teams

What makes collaboration effective in human–AI teams? How do roles shift, and who ultimately retains control? Leadership must be redefined to include “cyber-creative” decision-making processes, while maintaining human oversight through human-in-the-loop (HITL) approaches that ensure accountability and ethical responsibility.

Since AI can sometimes hinder idea generation and creative confidence, further research is needed to identify which AI tools are most suitable for creative contexts and how they can be used effectively. This includes systematic evaluation of impact, as well as the quantity and quality of creative outputs, with transparency, trust, and engagement at the core.

It is also essential to understand how humans can act to maximize synergy with AI agents. For instance, perceptions of AI—as a partner, assistant, advisor, or autonomous agent—may influence group dynamics and creative outcomes. Strategies that strengthen AI literacy could enable human participants to engage more effectively with these tools and critically assess their contributions.

The ultimate goal is to move beyond static role assignments and toward dynamic interaction between human and artificial agents, ensuring that each contributes optimally to the creative process.

6) Evaluating creative outputs

Are AI-assisted works less authentic? Or more efficient? It is important to explore how human audiences perceive AI-generated, human-generated, and hybrid cyber-creative outputs, while designing studies that minimize perception bias (which is often particularly severe toward AI-generated content).

We should investigate whether cyber-creative outputs can achieve authenticity—a quality defined by self-expression, lived experience, and intrinsic values, typically associated with human creativity.

New criteria and tools may be needed to distinguish between human and AI creativity. This could involve training AI systems on works recognized for their authenticity, as well as developing strategies to prevent the dominance of homogenized AI-generated content.

This implies designing algorithms that promote diversity, calibrating training data, and balancing human- and AI-generated content to support culturally rich and varied creative production.

Another key issue concerns transparency: how disclosure of AI involvement in creative processes affects trust and acceptance. Practices may be developed to clearly attribute authorship and specify the degree of AI contribution, positioning AI as a collaborative partner rather than a substitute for human creativity.

7) Exploring different domains

This type of innovation requires the development of frameworks that account for the specificities of individual disciplines, while also identifying universal principles that can guide more effective integration of technical and creative expertise across traditionally separate fields such as art, science, design, education, and innovation.

Rather than accepting conventional disciplinary boundaries, researchers should question whether these distinctions remain meaningful in an era of increasing integration.

This involves examining how cyber-creativity may bridge disciplinary divides, creating hybrid spaces where technical, creative, and other forms of expertise can be combined to address complex challenges that transcend conventional domain boundaries.

The focus thus becomes identifying barriers to interdisciplinary collaboration in cyber-creativity, including skill gaps, cultural incompatibilities, and structural constraints within organizations.

8) Transforming education

The educational domain, one of the most sensitive to AI’s impact, should emphasize support for learning rather than substitution of human effort, ensuring that AI enhances creative learning and human agency.

Future classrooms should teach creative thinking, ethical reasoning, and effective human–AI collaboration strategies, with teachers shifting from content transmission to creative mentorship. Professional development programs could redefine the teacher’s role as a mentor and facilitator, while also supporting psychological well-being.

Clear guidelines are needed for appropriate AI use, ensuring that it expands rather than replaces students’ ideas and voices. One possible approach is to move beyond simple question–answer prompting toward structured prompts that engage humans and AI in dialogical interaction, using question-based feedback to enhance creative thinking rather than deliver direct answers.

It is also important to explore co-creation models that enable teachers to adopt AI as a supportive creative collaborator, strengthening confidence and self-efficacy.

Furthermore, assessing how AI integration in educational practices can enhance teachers’ cyber-creativity skills and improve learning outcomes remains a key area of inquiry.

Additional solutions include integrated career guidance systems that combine algorithmic insights with human mentoring, helping students and families navigate an evolving labor market.

9) Addressing ethical issues

Ethical concerns represent a critical priority that cannot be postponed until “after the fact.” Protecting creators is fundamental. Issues of ownership and authorship are receiving increasing attention, with emerging legal and transparency frameworks aimed at preventing unauthorized appropriation of creative styles and ensuring proper attribution.

Transparency in AI, supported by open-source initiatives and detailed explainability reports, is essential for building trust and safeguarding artistic integrity.

Bias and cultural representation remain urgent issues. The predominance of Western and WEIRD (Western, Educated, Industrialized, Rich, Democratic) datasets has led to outputs that often misrepresent or exclude non-Western artistic forms, with the risk of cultural homogenization. This can result in distorted interpretations of cultural nuances.

It is therefore necessary to regulate interactions between AI systems and human creative practices, designing real-time governance models and ethical control interfaces that integrate human judgment into generative processes. In doing so, it becomes possible to uphold ethical standards even as systems become increasingly autonomous.

10) Confronting the darker side to prevent it

Potential risks of AI include the emergence of creative monopolies, cultural homogenization, and loss of agency. Monitoring and mitigation frameworks should be implemented to detect and contain the uncontrolled generation of potentially harmful cyber-creative content, while preserving creative freedom.

The goal is to ensure forensic traceability and robust identity protection, keeping pace with the rapid evolution of generative AI technologies. Adaptive machine learning models should be developed to analyze linguistic, contextual, and behavioral patterns in order to detect cyber-creative phishing attempts.

Research may focus on integrating these models into broader security frameworks, strengthening digital security for all users. Advanced verification models should move toward network analysis, content evaluation, and human fact-checking.

Researchers should also explore methods to identify and counter AI-driven disinformation, including signaling systems such as blockchain and watermarking, while safeguarding freedom of expression and privacy.

---

Creativity is part of our evolution, but it is also a step we can actively shape: waiting carries risks, and reacting too late may lead to a loss of autonomy. The scientific community and policymakers must act with urgency and collaboration, prioritizing research that defines frameworks capable of guiding policies aligned with the values we seek to preserve. In doing so, it may be possible to democratize creativity, expand human potential through new cultural forms, and safeguard society from homogenization and loss of meaning.