Skip to Main Content

AI in ECU Research: Copyright & Authenticity

The intersection of artificial intelligence (AI) and copyright law is changing quickly. This raises important questions about who owns and protects content created by AI. As AI technology improves, it is essential to understand the legal and ethical issues related to its use in creative work.

 

 

  • Typically belongs to the person or organisation that supplied the original ideas and information
  • Alternatively, it can belong to the individual who creatively executed the instructions in the prompt (Lucchi, 2024)

 
  • Works generated by AI without any human creative contribution are not eligible for copyright protection (Blaszczyk et al., 2024)
  • Copyright protects only original works authored by humans, including those produced with AI assistance
  • It does not extend to works entirely generated by AI (Blaszczyk et al., 2024)
  • AI is not considered a person in ethical and legal terms
  • This distinction is crucial in copyright law for determining ownership rights
  • AI systems that replicate existing information without significant original content do not qualify for copyright protection (Lucchi, 2023)
  • If the creation process lacks human input, the final work may lack originality and might not be eligible for copyright protection
  • AI-generated data can sometimes be entirely fabricated or echo previously established information
  • Generative AI is continuously being trained to create original content
  • This raises the question of whether AI-generated creations could qualify for copyright protection in the future (Barqawi et al., 2024)

 

  • AI is a new technology, so it's unclear whether works created with AI will be protected by copyright

In Australia, a work can only get copyright protection if a human author has contributed “independent intellectual effort.” This means that if an AI generates a work with little human input, it may not be eligible for copyright protection.

  • If creators use an AI tool and make significant contributions to their work, that work is likely to be protected by copyright.

However, it is not fully defined what level of contribution is enough for protection.

For example, if a creator makes an image and uses an AI tool to edit it, that human input is probably enough for copyright protection. But it's still unclear if other contributions - like writing, adjusting the algorithm, or selecting and editing output - would also qualify for copyright protection. 

Read More: Arts Law Centre of Australia. (n.d.). Information sheet: Artificial Intelligence (AI) and Copyright

It is usually against the law to copy digital works without permission. If you want to make a digital copy of a work, including those found online, to train an AI, you need to get permission from the copyright holder. For example, if you create your own training dataset by collecting images from publicly available websites, you must ask for permission from the rights holders.

However, you might not need permission if certain exceptions apply. For example:

  • Fair dealing exceptions exist for research, study, criticism, review, news reporting, and parody or satire.
  • Some exceptions allow temporary copying for technical reasons, like when your computer makes a temporary copy of a video while you stream it from a licensed platform.

Read More:

Fair dealing umbrella. When does it apply?

  •        Using copyrighted text or images to train AI models that generate unique, non-derivative outputs
  •        AI analysing large datasets without distributing copyrighted content
  •        Summarising or referencing parts of a book rather than replicating it entirely
  •        AI-generated content does not replace the original work, nor does it affect its value

 

 

ECU's Framework for the Productive and Ethical Use of Artificial Intelligence is built upon five key Ethical Principles. These principles empower staff and students to make informed decisions about AI at all levels and across all areas of the organisation, allowing us to navigate our ethical engagement with AI collectively.

Benchmarking shows that widely accepted ethical principles for using AI match ECU's values. This connection strengthens the trust in ECU's AI Ethical Principles for everyone involved.

The Respect Value states: ECU is committed to protecting privacy, intellectual property, equity, diversity, and inclusion in our approach to Artificial Intelligence.