In recent years, Artificial Intelligence (AI) has become a hot topic, with discussions surrounding its use and associated risks intensifying. However, increasing attention is being paid to how AI acquires data and undergoes training. Recently, a Melbourne-based publisher requested permission from its authors to use their works for AI training, sparking crucial discussions about copyright and intellectual property.
Intellectual property protection encompasses patents, trademarks, copyrights, and designs, with copyright holding particular significance in the AI field. Legal experts highlight that copyright protects the expression of a work, not the idea itself. For instance, with literary works, artwork, books, and illustrations, copyright grants authors several rights, most importantly the right of reproduction. Ensuring non-infringement of copyright in AI-generated content is a key issue.
Image Source Note: Image generated by AI, licensed through Midjourney.
During AI training, the AI needs to reproduce the content of the original work. Using a work for AI training without the author's permission may constitute copyright infringement. Experts argue that AI-generated outputs typically draw from multiple sources and therefore, in many cases, are not considered substantial reproductions of the original work. This is not the first instance of AI companies facing lawsuits for unauthorized material use; numerous legal disputes have erupted in various countries.
Furthermore, AI can produce biases and misinformation when processing data; this "fake news" can have severe consequences for users. For example, lawyers and students have been penalized for citing fabricated cases from AI-generated content. The academic community also calls for appropriate credit to be given to authors and for their rights to be protected.
Current copyright laws lack specific legislation addressing AI, relying instead on existing legal frameworks. Experts recommend that authors first verify if their works are being used for AI training and advocate for increased transparency in datasets. Moreover, a more robust legal framework is needed to ensure the accuracy and compliance of data usage.
With rapid technological advancements and upcoming elections, experts say the government needs to take urgent action to protect creators' rights.
