England will examine training data of artificial intelligence models

Copyright issue of artificial intelligence will be addressed

AI models are trained on massive data sets of text, images and other content. However, in this process, some artists and publishers are expressing their discomfort with their content being used in this way without their consent. Google ve OpenAI Claiming that large technology companies such as Collect their content without permission, these groups demand that their rights be protected.

Tensions in this area have increased especially in the past year. The New York Times‘of Microsoft ve OpenAI’ye came to the fore with the lawsuit he filed against him. The newspaper claimed that copyrights were violated and intellectual property was misused during the training of artificial intelligence models. OpenAI rejected these accusations and argued that the use of open web data should be considered within the scope of “fair use” and stated that they offer rights holders an “opt-out” option to remove their content from their systems.

The UK government will consider the following proposals during the consultation process:

• Copyright exception: Making an exception to copyright law for artificial intelligence training for commercial purposes; but granting rights holders the authority to control the use of their content.
• Licensing and pricing: Creative content owners have access to licensing and payment mechanisms so that their works can be used by AI models.
• Transparency requirements: AI developers to share more information about the datasets used to train their models and offer clarity to content owners about this process.

This transparency proposal could be controversial for tech companies. Due to commercial competition, companies avoid providing information about the data they use to train their algorithms as much as possible. Meanwhile, in the USA, regulations on artificial intelligence are becoming complicated due to the strong lobbying activities of giants such as Amazon, Microsoft and Google. However, the United Kingdom managed to establish a solid ground in this field with the Data Protection Act of 2018.

As artificial intelligence models develop rapidly, these systems are no longer limited to text only. It is gaining “multi-modal” capabilities such as image and video production. Last week, OpenAI rolled out its new video production model, Sora, internationally. Users can create high-resolution video clips with text input. With the advent of multimodality, the problem of copyright infringement also becomes more salient.

Share via Email
This is titled mail it to your friend.

This news our mobile application Download using
You can read it whenever you want (even offline):