Disclosure Requirements for Copyrighted Material Use for OpenAI

As artificial intelligence continues to advance, disclosure requirements for copyrighted material used in AI training have discussed in the European Union. Companies developing content-generating tools, such as ChatGPT, Dall-E, and Midjourney have to reveal their use of copyrighted material in training their systems.

disclosure requirements for copyrighted material

The AI Act and Copyright Transparency

In the last two weeks, the requirement for transparency concerning copyrighted material usage was added to the Artificial Intelligence Regulation (AI Act). Some European Parliament members initially advocated for banning the use of copyrighted material. This gives a more lenient proposal focusing on transparency ultimately prevailed.

Implications for AI Companies | Disclosure Requirements for Copyrighted Material

Generative AI systems rely on extensive data training to understand various writing styles, acquire information, and learn diverse artistic techniques. Companies like OpenAI are historically secretive about the data to train their software. However, if European legislation enacted, these companies will need to disclose the sources of their content.

Potential Consequences After Disclosure Requirements for Copyrighted Material

  • Increased copyright infringement lawsuits
  • Stricter Regulations on AI training data
  • A shift in the AI development landscape

The issue of copyright is already a significant concern in the AI space, particularly among artists and photographers. For instance, Getty Images is currently suing Stability AI, whose AI model Stable Diffusion was found to generate images with the Getty watermark, suggesting misuse of copyrighted material.

Moreover, three artists filed a class action against Stability AI, Midjourney, and DeviantArt. This alleges that the developers violated the rights of millions of artists by using 5 billion images without consent.

In a separate case, Microsoft and its subsidiary GitHub are facing a lawsuit over the Copilot tool. Copilot was using open-source code protected by licenses that require author attribution in derivative works. However, the tool generates long scripts from these works without providing credit, potentially violating copyright law.

More articles from us:

Conclusion

As the European Union considers implementing disclosure requirements for copyrighted material used in AI training, companies developing AI systems may face increased scrutiny and legal challenges. Moreover, ensuring transparency and compliance with copyright regulations will become increasingly important for AI developers as the industry continues to evolve.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top