Introduction
In a world where artificial intelligence (AI) is developing at breathtaking speed, the question of intellectual property (IP) arises with renewed urgency. Recent advances in AI have not only transformed entire industries but have also given rise to entirely new creations, often produced without direct human intervention. This reality raises fundamental questions about how such creations should be protected by law, if indeed they should be at all.
In January 2025, the United States Copyright Office published Part 2 of its report on AI: the outputs of generative AI are protectable only where a human author has determined sufficient expressive elements; a mere prompt is not enough. This publication sparked a global debate on the need for new legal frameworks to address this new frontier. In parallel, the European Union began enforcing the AI Act from 2 August 2026, a framework based on four levels of risk. These developments clearly show that AI and IP have become a new arena of conflict, requiring ongoing dialogue and the rapid adaptation of existing laws.
Context/History
The relationship between AI and intellectual property has gone through key stages over the years. Since the emergence of AI in the 1950s, nearly 340,000 AI-related inventions have been the subject of patent applications, and more than half have been published since 2013 (WIPO, 2019). However, it was only with the advent of neural networks and machine learning algorithms in the 2010s that the issue became central. The rise of natural language models such as OpenAI's GPT illustrated the creative potential of AI, posing unprecedented challenges for regulators and legal experts.
Some of the first legal battles over AI and IP focused on whether inventions made by AI could be patented. The debate remains open depending on the jurisdiction, and concerns the machine less than inventiveness and human contribution.
Applications/Use Cases
The applications of AI in content creation are vast and varied. For example, companies such as OpenAI (with GPT), Stability AI (with Stable Diffusion) and Midjourney have developed models capable of generating text and images, used in marketing, content writing and even artistic creation. These tools make it possible to produce high-quality content quickly, but they also raise the question of who owns the works generated in this way.
In healthcare, Google DeepMind used its AlphaFold system to make significant discoveries about protein folding, an advance that could transform biomedical research. However, these discoveries raise the question of intellectual property: who owns the results generated by a machine?
Technologies/Methods
The underlying technologies that enable these advances in AI-driven creation mainly include machine learning and deep neural networks. These systems learn from large amounts of data in order to imitate or surpass human capabilities in certain tasks. Recent language models have a number of parameters that their publishers do not always disclose, and they produce text that a reader cannot always distinguish from text written by a human.
AI content generation techniques also raise questions of transparency and traceability. Systems must be designed so that they can be audited and so that it is possible to understand how decisions are made, which is essential for addressing intellectual property issues.
Challenges/Limitations
The main challenge of AI with regard to intellectual property lies in defining human contribution. AI systems can produce works of art, music and even technical inventions, but the question remains whether these outputs can be protected by copyright or patents if no human intervention can be identified.
There are also concerns about ethics and liability. Who is responsible when an AI infringes intellectual property? Companies must navigate a complex legal landscape to ensure that they do not infringe existing rights while innovating.
These questions of governance and data ownership are not theoretical for companies that deploy AI in direct contact with their users. At Koaee, we design multilingual conversational assistants for businesses, available in sixteen languages, with servers located in Europe and no conversations retained. The models we use — Mistral (Voxtral) and OpenAI (GPT) — are part of this evolving technological landscape, which these debates on AI and intellectual property are helping to shape.
Outlook
Going forward, we can expect intellectual property laws to continue evolving in response to technological advances. New legal frameworks will likely be needed to address the specific characteristics of AI-generated works. In addition, international standards could be developed to ensure a consistent approach across jurisdictions.
The concept of collaborative models, in which AI and humans work together, could also influence how intellectual property is managed, with an emphasis on co-creation and the recognition of shared contributions.
Conclusion
As AI continues to transform creativity and innovation, the question of intellectual property is becoming increasingly pressing. Regulators, businesses and researchers must work together to create frameworks that protect rights while fostering innovation. The way forward will require a balance between regulation and flexibility, enabling AI to reach its full potential without compromising intellectual property rights.
Source: Eurostat, survey “Artificial intelligence by size class of enterprise” (isoc_eb_ai) — businesses with 10 to 249 employees, France. Retrieved via API on 30 August 2026. The survey publishes neither 2022 nor 2026: the series ends at the latest available year.
