The EU AI Act does not provide explicit provisions regarding the intellectual property rights of AI-created work. Here’s where the legal perspectives from various parts come in:
Standardisation should play a key role to provide technical solutions to providers to ensure compliance with this Regulation. (Recital 61)
This recital suggests that providers of AI technology are expected to comply with defined standards and regulations. Consequently, if rules on intellectual property rights for AI-generated creations were to be defined, AI system providers would be obliged to comply with them. This doesn’t entirely answer the question but indicates the potential direction of responsibility under such circumstances.
A Union legal framework laying down harmonised rules on artificial intelligence is therefore needed to foster the development, use and uptake of artificial intelligence in the internal market that at the same time meets a high level of protection of public interests, such as health and safety, protection of fundamental rights, democracy and rule of law and the environment, as recognised and protected by Union law. (Recital 5)
This recital suggests the existence of a harmonized legal framework within the EU to protect fundamental rights and public interests, in relation to AI usage. It does not directly mention intellectual property rights but implies the need for legal protection of interests within the scope of AI operation. The mention of “fundamental rights” might potentially include such rights as intellectual property.
Now, it’s important to mention that while Articles 4, 12, 28, and 70 and Annex IV of the Act were all examined, none of them provided any specific directives on the question of intellectual property rights of AI-created works.
The lack of precise information underscores the complexity of the issue. Legal ownership of creations by AI is a topic of ongoing debates around the world, with no unanimous decision. The discussion expands beyond the framework of AI and into broader fields such as intellectual property law and copyright. As the AI Act evolves, future amendments may clarify the EU’s position on this matter.
In conclusion, the Act does not specify who holds intellectual property rights for AI-generated creations. Therefore, other legislative acts, judicial precedents, or future provisions in the AI Act should be explored to answer the question adequately. Consulting with a legal professional specialized in AI and intellectual property law may also provide further insights.
Based on the analysis of the AI Act, none of the referenced Articles and Titles directly address the question of intellectual property rights for AI-made creations. The Act instead focuses on the responsibilities and transparency obligations of operators and providers in the AI value chain. For instance:
“The operator shall put in place adequate measures to minimise the risks that the AI system poses to health and safety or fundamental rights, commensurate with the nature of the risk as well as with the probability and the severity of the potential damage.” (Article 28 of the AI Act)
This implies that while obligations are placed on operators, the Act doesn’t explicitly attribute ownership or rights related to creations of AI systems to any particular entity.
The same applies to Article 4 on the general principles for trustworthy AI, Article 5 on prohibited AI practices, and Title IX on codes of conduct. These elements of the Act provide guidance on the ethical use and regulation of AI practices but do not dictate intellectual-property rights.
Given that the EU AI Act does not directly assign intellectual property rights for AI creations, it is necessary to consult relevant IP law for works created using AI, including EU copyright law, EU patent law, and the recommendations of the World Intellectual Property Organization (WIPO). This research would allow for a more comprehensive understanding and answer to the legal question.
It appears the AI Act doesn’t specifically address the question of who owns intellectual property rights of an AI-made creation. However, we can infer some implications from several relevant provisions in the Act.
”For the purposes of this Regulation, the following definitions apply:
(1) ‘Artificial Intelligence System’ (AI system) means software that is developed with one or more of the techniques and approaches listed in Annex I and that can, for a given set of human-defined objectives, generate outputs such as content, predictions, recommendations, or decisions influencing the environments they interact with.” (Article 3)
This definition gives us a clear understanding of what constitutes an AI system. The AI Act characterizes an AI system as a tool or instrument capable of generating outputs such as content but doesn’t explicitly attribute ownership of these outputs to anyone. It’s software used to fulfill human-defined objectives.
While the AI Act outlines a range of obligations and restrictions for AI systems and their use, none specifically relate to the ownership of IP rights of AI outputs:
“AI systems should be designed and developed in a way that respects fundamental rights, the rule of law, and the principles acknowledged in particular by the Charter, including human dignity, freedom, democracy, equality, the rule of law, and respect for human rights.” (Article 4)
The AI Act does provide some guidance on issues of transparency, trustworthiness, and accountability, but again, there are no direct references to IP rights or the ownership of AI-generated creations.
”AI systems that are subject to third-party conformity assessment… shall have a risk management system in place, which shall be proportionate to the nature and function of the AI system.” (Article 9)
Risk management provisions are focused on ensuring safety and minimizing risks so it’s not concerned with the IP rights of AI outputs.
Given that the AI Act doesn’t provide a direct answer to this question, the legal question of who owns the intellectual property rights of an AI-made creation would likely be determined by other existing legal frameworks, notably EU intellectual property law, and national legislation where applicable.
Please consult an IP law specialist for more comprehensive advice on this topic.
(2b. For the purposes of this Article, trade secrets shall be preserved and shall only be disclosed provided that all specific necessary measures pursuant to Directive (EU) 2016/943 are taken in advance to preserve their confidentiality, in particular with respect to third parties. Where necessary, appropriate technical and organizational arrangements can be agreed to protect intellectual property rights or trade secrets.) (Article 28)
This quote emphasizes the preservation of trade secrets when using AI systems, stating that all necessary measures be taken in advance to ensure their confidentiality. While it doesn’t directly answer your question, it implies a level of control and protection over information considered intellectual property. The specifics of ownership may rely on individual agreements or jurisdiction-specific intellectual property laws.
The developers of free and open-source AI components should not be mandated under this Regulation to comply with requirements targeting the AI value chain and, in particular, not towards the provider that has used that free and open-source AI component. (Recital 12c)
This statement indicates that developers of open-source AI components are not bound by certain requirements of the AI value chain. While this doesn’t directly answer your question, it implies that users of open-source components might not have commitments towards the original developer, although this could be influenced by licenses or other agreements.
A number of legally binding rules at European, national and international level already apply or are relevant to AI systems today, including but not limited to EU primary law (the Treaties of the European Union and its Charter of Fundamental Rights), EU secondary law (such as the General Data Protection Regulation, the Product Liability Directive, the Regulation on the Free Flow of Non-Personal Data, anti-discrimination Directives, consumer law and Safety and Health at Work Directives), the UN Human Rights treaties and the Council of Europe conventions (such as the European Convention on Human Rights), and national law. (Recital 41a)
The AI systems today are subject to numerous legal principles at different levels. It indicates that other legal frameworks, such as EU Treaties and Charter, data protection regulations, product liability directive, and laws related to consumer law, among others, may also have a bearing on your question.
the methods and steps performed for the development of the AI system, including, where relevant, recourse to pre-trained systems or tools provided by third parties and how these have been used, integrated or modified by the provider (Annex IV-2(a))
This provision points out that there could be multiple entities involved in an AI system’s development, tipping us off to the possibility of shared intellectual property ownership. This would largely depend on agreements in place and the nature of individual contributions.
A description of the architecture, design specifications, algorithms and the data structures including a decomposition of its components and interfaces, how they relate to one another and how they provide for the overall processing or logic of the AI system (Annex IV-2(b))
This requirement focuses on the design and architecture of AI systems. Depending on the nature of these features, they could be subject to IP protection, with ownership rights being largely determined by who designed and implemented said features.
The AI Act does not directly articulate the ownership of intellectual property rights over creations made by an AI. Who retains these rights might depend on terms set out in individual agreements, the nature of the intellectual property, and laws specific to the jurisdiction in question. For a definitive answer, legal advice from an expert in intellectual property law, particularly in the context of AI, may be necessary.