EU AI Act and Its First Code-of-Practice Draft
The European Union has recently made strides in the realm of artificial intelligence with the introduction of the AI Act. This legislation aims to regulate the use and deployment of AI technologies throughout the EU, ensuring that ethical standards are maintained in the development and implementation of AI systems.
Overview of the AI Act
This year, the music industry celebrated the passing of the AI Act. However, the focus has shifted towards the effective implementation of this legislation, which has been a point of contention amongst various creative-industry bodies. These organizations are advocating for a more meaningful application of the AI Act across the EU, emphasizing the need to address the concerns surrounding copyright and data usage.
First Draft of the General-Purpose AI Code of Practice
In mid-November 2024, a first draft of the “General-Purpose AI Code of Practice” was released. This document is the initial step in a series of four drafting rounds planned to refine the legislation, with the finalization expected by April 2025. The code primarily addresses general-purpose AI models, which are designed to perform multiple tasks and are widely used across various sectors.
Key Provisions on Copyright
A notable aspect of the draft is its emphasis on copyright issues. Companies that sign the code are required to acknowledge that any use of copyright-protected content necessitates the authorization of the respective rightsholder(s), unless specific exceptions apply. This clause underscores the importance of intellectual property rights in the context of AI.
Furthermore, the draft mandates that AI firms conduct “reasonable copyright due diligence” before entering into contracts related to the use of data sets for developing AI models. This provision aims to ensure that all parties involved in the data usage are aware of and comply with existing copyright laws, thereby protecting the rights of content creators.
Transparency and Accountability Measures
Another significant requirement of the code is the obligation for signatories to maintain and provide up-to-date information regarding the data sources used for training, testing, and validation of AI models. They must also disclose authorizations to access and utilize protected content. This transparency measure is expected to be positively received by rightsholders and is aimed at fostering accountability within the AI sector.
Future Outlook
While the draft is a step forward, it is crucial to note that it is still in its preliminary stages. The document will undergo extensive scrutiny from nearly 1,000 stakeholders, including technology firms, creative industry representatives, and policymakers. Their feedback will play a vital role in shaping the final version of the code.
The ongoing discussions and debates surrounding the AI Act highlight the intricate balance that needs to be achieved between innovation in technology and the protection of intellectual property rights. As the drafting process continues, stakeholders will eagerly await the outcomes that will ultimately define the landscape of AI regulation in the European Union.
